RE: PgAccess 0.97 - final preview

Constantin Teodorescu wrote:
>
> Hiroki Kataoka wrote:
> >
> > I think that better solution is to separate a source file of pgaccess to
> > main code and message catalog files. And message catalog files should be
> > supplied for each of all supported languages. The pgaccess main code
> > includes only english messages as default language, and
> supports to load one
> > message catalogue file at startup.
>
> That's exactly what I intend to do in the next version.
> But in order to grab together more translations and for the easy of use
> I choose to keep for the moment the Messages array in the main file.
> There will be separate *.lang files that will be used for different
> languages. Also , PgAccess will be sliced into separate files for each
> module.
>
> But related to japanese messages I have a question. If I will move the
> definition for Messages in a separate file, why would it help it? What
> "japanese Tcl" are you talking?
>
> Have you tried Tcl/Tk 8.1 or 8.2 beta? They (scriptics) are saying that
> it support Unicode. Maybe that's what you should use.


I understood your plan.

The japanese Tcl has some new functions; kanji encoding auto detection,
kanji encoding conversion, etc.. These encoding functions are needed for
japanese users, because japanese kanji texts are represented in many
encodings (JIS, ShiftJIS or EUC). But one text is not represented in more
than one encoding, because the kanji encoding auto detection is impossible
when a text is represented in more than one encoding.

I know that latest Tcl version is 8.1 and this version supports Unicode. But
using this version on japanese environment, we have a obstacle when using
kanji input method. I think that Tcl 8.1 can not use practically in japanese
environment for the time being. By the way, almost all japanese users are
using japanese enhanced Tcl7.6 and Tcl8.0.x.

Let us suppose that japanese users use Tcl8.1. Because Tcl8.1 assumes that
Tcl source file is written in running operating system's encoding, we can
not write source file in Mixed languages, either.

If one source file will include one language only, then both japanese Tcl
and international Tcl8.1 will load the Tcl source file successfully.

Thank you.
Hiroki Kataoka