Re: [ft-devel] Question about charset.

[suzuki toshiya]

It's my shame that I'm too late to reply to the question
including CJK keyword.

The codepage info in the fonts designed for Microsoft
platform is basically used to distinguish the market
that the typeface designer assumed, not to analyze the
charset in the font. For example, China mainland, Japan,
Korea and Taiwan have different type-design preference.

In fact, although TrueType and OpenType specs use a
name of "GB-2312", the regional charset standard for
P.R.China, but, most fonts for PRC market include
wider charset, like, GBK.

Regards,
mpsuzuki

Johnson Y. Yan wrote:
> Thanks, Werner.
>
> Johnson Y. Yan
>
>
>
> 发件人： Werner LEMBERG
> 发送时间： 2013-04-19 12:27:57
> 收件人： yinsen_yan
> 抄送： freetype-devel
> 主题： Re: [ft-devel] Question about charset.
>
> > I mean that it likes 
> >
> > #define FT_WinFNT_ID_CP1252    0
> > #define FT_WinFNT_ID_DEFAULT   1
> > #define FT_WinFNT_ID_SYMBOL    2
> > [...]
> >
> > defined in ftwinfnt.h, can these work for other type fonts, type1,
> > CFF and so on.
> No.  In general, FreeType only handles encodings, not charsets.
> Additionally, information on the character sets within a font is
> *very* unreliable.  For example, the Unicode block bits in the `OS/2'
> table of TrueType fonts are set by font vendors rather arbitrarily.
> They are just a possible iteration to the question: `Is this font
> suitable for displaying this range?', and there is absolutely no
> guarantee that all characters of this block are really available in
> the font.
>     Werner
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Freetype-devel mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/freetype-devel