Re: UTF to CID-Keyed

From: Jungshik Shin (jshin@pantheon.yale.edu)
Date: Tue Apr 11 2000 - 17:52:36 EDT


On Tue, 11 Apr 2000, gunnarb wrote:

> Jungshik wrote
>
> > Did you try files in/under
> >
> > ftp://ftp.oreilly.com/pub/examples/nutshell/cjkv/adobe
> >
> > and references cited in README ?
> >
> Hi Jungshik, thank you for responding, and yes, I've got Ken Lunde's
> resources, at least the CMaps and CID-Keyed font character collections
> along with Adobe's CCDs and Acrobat 4, so I'm able to happily generate
> glyphs in Chinese (GB-K, Big-5 and HK-supplement), Korean and Japanese
> -- three languages I don't understand. The next step is to take the
> information from Oracle 8, where it's stored presently in UTF16 and
> translate that into something the CMap will recognize. Perhaps I've
> missed part of the README, but I'm not seeing that critical step.

> The only other possiblity I can see is Mr. Lunde's program JConv ,which
> converts the Japanese encoding in text files, He has made the ANSI C
> source code and a Perl version are also avialable, ostensibly so that
> programmers may use the algorithms to convert among other Japanese
> encodings. Perhpas it can be modified for use with Chinese as well ... ?

 If what you need is a covnerter among various Chinese legacy encodings
and UTF-16, you can use native2ascii included in JDK assuming
your data only uses UCS-2 part of UTF-16. Or you can write a simple
program to convert UTF-16 to UTF-8 and use native2ascii.
(or Oracle should support output in the encoding of your choice.
another alternative is using iconv() in modern I18Nized
Unix or equivalent offered by your OS)

  % utf16toutf8 input.utf16 | native2ascii -encoding UTF8 | \
    native2ascii -reverse -encoding Big5 > output.big5

  But, I guess this is only half of the job in creating CMap files for
Big5, GB-K and so fonts if you have different glyph IDs than used by
Adobe. As you can see by looking at CMap files, it's not just a straight
mapping, but it shouldn't be hard to figure out how to make one with a
simple script. (I used to know this, but forgot....)

  Well, you'd get the best and speediest answer by writing
directly to Ken Lunde :-).

    Jungshik Shin



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:01 EDT