Re: Pango Unicode coverage

[James Su <suzhe gnuchina org>]

Hi,
There is a TTF font released by Founder Corp. China, which includes more
than 70000 characters. Please visit:

http://www.founder.com.cn/fontweb/chanpinzl/CP_chaoda.htm

and

http://www.founder.com.cn/fontweb/chanpinzl/FA_chaoda.htm

to get the detailed information for that font (It's in Chinese).

The font uses *surrogate* space in Unicode to encode those characters. I
think it should use UTF-16 as its internal encoding. According the
article, Windows 2000 has support it.

This font is included in Office XP (at least in its Chinese Version).
The font size is 40M bytes. Here is the ftdump output for this font:

    font name table entries

------------------------------------------------------------------------------
    Simsun (Founder Extended) - Version 1.00
    PostScript name: FZSY--SURROGATE-0
    Copyright(c) Founder Corporation.2000
    By Founder Corporation. 'ąc

------------------------------------------------------------------------------
    character map encodings

------------------------------------------------------------------------------
    There are 3 encodings:
    encoding 0: Apple Roman
    encoding 1: Windows Unicode
    encoding 2: Windows Unknown value 10

------------------------------------------------------------------------------
    ftxcmap test

------------------------------------------------------------------------------
    There are 3 encodings:
    encoding 0:
    first: glyph index 1, character code 0x0
    next: glyph index 0, character code 0x1
    last: glyph index 191, character code 0xff
    encoding 1:
    first: glyph index 1, character code 0x20
    next: glyph index 2, character code 0x21
    last: glyph index 0, character code 0xffff
    encoding 2:
    first: glyph index 0, character code 0xffffffff
    next: glyph index 0, character code 0xffffffff
    last: glyph index 0, character code 0xffffffff

------------------------------------------------------------------------------
    Memory footprint statistics:

------------------------------------------------------------------------------
    525981 Bytes ( 514 kByte): face object
    7835 Bytes ( 8 kByte): glyph object
    1356 Bytes ( 2 kByte): instance object
    15639 Bytes ( 16 kByte): exec. context object

------------------------------------------------------------------------------
    550811 Bytes ( 538 kByte): total memory usage


James Su

Roozbeh Pournader wrote:

>On 26 Dec 2001, Owen Taylor wrote:
>
>> * The unicode character property code in Glib needs some
>>   work to handle properties for characters not in the
>>   BMP. (The interfaces are fine, but the internal data
>>   structures are limited.)
>>
>
>We recently redid FriBidi's informational data structures to support
>non-BMP characters. Do you think it will be hard to do for Glib? If not, I
>can put some time on it (mainly for demonstration purposes, since I love
>to refer people to pango).
>
>>   (Can the TTF format handle Unicode charmaps with characters 
outside the
>>   BMP? I don't think so...)
>>
>
>I've heard that it does, but did not get into the details. As far as I
>know, there is a CJK font in WinXP that does this. (There is only a limit
>of 65536 glyphs per TrueType font, which should not be a problem.)
>
>roozbeh
>
>_______________________________________________
>gtk-i18n-list mailing list
>gtk-i18n-list gnome org
>http://mail.gnome.org/mailman/listinfo/gtk-i18n-list
>
>