Subject: Re: adding CP1251 support in AbiWord
From: Alexander Shopov (al_shopov@web.bg)
Date: Fri Oct 26 2001 - 03:08:12 CDT
Hi Tomas,
> I gather you need the ttftool to support cp1251, that is easy to do, I
> will get that done in next couple of days, I only need you to tell me
> what Unicode character represents character 0x98 in cp1251, for
> some reason it is missing in my i18n maps.
I am investigating ...
BTW Pablo Saratxaga <pablo@mandrakesoft.com> has implemented similar
things for the program ttmkfdir.
Here are links to the source and binaries:
ftp://ftp.lip6.fr/pub/linux/distributions/mandrake-devel/cooker/SRPMS/freetype-1.3.1-12mdk.src.rpm
ftp://ftp.lip6.fr/pub/linux/distributions/mandrake-devel/cooker/i586/Mandrake/RPMS/freetype-tools-1.3.1-12mdk.i586.rpm
(these are links to cooker packages, you may want to try the directories
if Mandrake upgrades freestype between the moment I write and the moment
you see the letter)
Still, I will look into the 0x98 in cp1251 problem and send you some
answer with a link to some authoritative web page.
>
> If the only difference between the font encoding and the encoding
> part of LANG is capitalsation, it should not cause any problems,
> just use the version that is returned by LANG as a name for the
No, this is not what I meant.
I read the http://www.abisource.com/lxr/source/abi/UnixLocales.txt three
more times and I see my mistakes.
Sorry for raising false alarms. ;-(
still - what I meant is as follows:
<explanation_of_stupidity owner=al_shopov>
What I meant is: AbiWord handles encodings other than iso8859-1, as a
subdirectory in its font directory.
Now compare:
For KOI8-R
The proper name of the locale is for example: ru_RU.KOI8-R
The proper name of the fonts dir is
<AbiRoot>/share/AbiSuite/fonts/KOI8-R
And the encoding of the font (the last two fields of its X name) is
-*-*-*-*-*-*-*-*-*-*-*-*-koi8-r
Now for cp1251
The proper name of a locale is bg_BG.CP1251
It seems to me the proper name of the fonts dir is
<AbiRoot>/share/AbiSuite/fonts/CP1251
*BUT* what we call encoding of the font (the last two fields of its X
name) is -*-*-*-*-*-*-*-*-*-*-*-*-microsoft-cp1251
</explanation_of_stupidity>
Now I see that the situation is the same for UTF-8 and ISO-8859-*
locales - the name of the encodings of charsets and fonts differs there
as well. In fact the locales and fonts names have nothing to do with
each other. Tomas, can you please stress it stronger in the file
http://www.abisource.com/lxr/source/abi/UnixLocales.txt - that it is the
encoding of the locale that is important (lines 38-39), not the font!
(this is for exceptionally dumb people as my not-so-humble self)
Best regards:
al_shopov
This archive was generated by hypermail 2b25 : Thu Oct 25 2001 - 02:09:09 CDT