On Sun, 28 Jan 2001, Tomas Frydrych wrote:
Hi,
In the following fragment of the patch:
+ if(!strcmp(enc, "UTF-8"))
+ m_bIsUnicodeLocale = UT_TRUE;
+ else
+ m_bIsUnicodeLocale = UT_FALSE;
I think we also need to check whether the locale name is "UTF8" too - as I
recall new glibc removes all non-alpha non-digit chars from locale name (e.g.
KOI8-R becomes KOI8R).
At least this fix won't hurt.
Also, text becomes misformatted after this patch (if this matters at all).
>
> The attached patch enables the use of Unicode fonts for utf-8
> locale. Utf-8 locales need to be able to access arbitrary Unicode
> characters in a Unicode font; further the utf-8 locale differs from
> other locales, such as the iso8859-* set, in that the utf-8 coding is
> not used by its native fonts which are coded using UCS-2.
> Consequently, neither the single-byte mechanism used by the ISO-
> 8859-* locales, nor the multibyte method we use for CJK can
> handle utf-8 locale. This patch implements the necessary changes,
> the behaviour for all other supported locales is uneffected.
>
> Tomas
>
Best regards,
-Vlad
