Re: [Groff] Problems with unwanted unicode.

[Tomohiro KUBOTA]

Hi,

At Fri, 13 Jul 2001 04:14:21 +0200 (CEST),
Werner LEMBERG <address@hidden> wrote:

> A better temporary solution is probably to add a single flag like -D1
> and -D0 to activate interpretation of double-byte characters
> independently of the output device.  Additionally, an environment
> variable like GROFF_DBCS might help to permanently activate it for
> Japanese users.

I wrote a similar patch, though it is against Debian version of groff.

    http://www.debian.or.jp/~kubota/packages/groff/

I added a new (tentative) command option "--eucmode".  (I hesitated
to use one-character option name.)  However, I think this solution
is not very good.


> This `poor' design is due to the age of troff.  BTW, TeX has similar
> flaws, and the designer of LaTeX 2e had a lot of work to fix it.

I understand.  It is relatively in the early stage of Unix development.
In such an age, it is natural that the developers didn't think about
internationalization.


> I'm going to fix this for groff 2.0 by implementing Unicode,
> separating input and output encoding, but it won't be a quick
> development process...

Then, could you think about tentatively integrate Japanese patch
into official version of Groff?  The Japanese mode would be enabled 
when the roff source has a line to specify EUC mode.  (Do you remember
that the preprocessor will read the roff source and understand MIME
style and Emacs style encoding specification?  I am thinking about
that troff core reads the encoding specifier and enable Japanese
mode for Japanese encoding.  I think Japanese mode is also useful
for Chinese.  Though Korean (EUC-KR) is also a multibyte encoding,
they don't need Japanese mode since they use whitespace between 
words.)

---
Tomohiro KUBOTA <address@hidden>
http://www.debian.or.jp/~kubota/
"Introduction to I18N"  http://www.debian.org/doc/manuals/intro-i18n/