[Glade-users] About Japanese characters code in build operation

[damon ximian com (Damon Chaplin)]

> atuhiko wrote:
>
> Hello.
>
> There is a question about Japanese characters code in build operation.
> I carried glade out with the following environment at first.
>  RedHat 6.2
>  glib 1.2.6-3
>  gtk+ 1.2.6-7.
>  glade 0.5.5-4
> Japanese characters were kept with the EUC-JP code in the glade file.
> It was kept with the EUC-JP code in the interface-creation-functions file
>  at the time of build operation as well.
>
> But, a problem occurred when I carried glade out with the following environment.
>  RedHat 7.1
>  glib 1.2.9-1
>  gtk 1.2.9-4.
>  glade 0.5.9-5
>
> Japanese characters were kept with the unclear code in the interface-creation-functions file
>  though Japanese characters were kept with the EUC-JP code in the glade file.
> There is not even which of EUC-JP, SJIS, JIS, UCS-2, UTF-8 in it.
> Tell me which Japanese characters code to output .
> Incidentally, as an example.
> zero of two bytes character  "��" = 0x5c 0x32 0x34 0x33 0x5c 0x32 0x36 0x30
> The capital letter A of two bytes character "��" = 0x5c 0x32 0x34 0x33 0x5c 0x33 0x30 0x31
> The small letter a of two bytes character "��" = 0x5c 0x32 0x34 0x33 0x5c 0x33 0x34 0x31
> The first character of Japanese hiragana "��" = 0x5c 0x32 0x34 0x34 0x5c 0x32 0x34 0x32

I'm afraid I don't know how Glade should handle different character encodings,
especially 2-byte encodings for Chinese/Japanese/Korean.

Currently Glade has no knowledge of character sets or encodings. It just
saves everything as strings, as returned by gtk_entry_get_text() or whatever.
I thought that would work as long as the character set used on the target
machine matched that used while creating the interface. But I'm not sure.


The easiest thing to do at the moment is probably to use English in Glade to
create the user interface, and then use gettext to translate it into other
languages later. Though I realize this isn't ideal.

Hopefully things will improve in GTK+ 2.0, when we will probably move to UTF-8
for everything.

If anyone has ideas for what we should be doing to fix this problem let me know.

Damon