Subject: Re: International support in Win98
From: Vlad Harchev (hvv@hippo.ru)
Date: Thu Dec 21 2000 - 14:54:21 CST
On Thu, 21 Dec 2000, Lukas Pietsch wrote:
> Thanks to Vlad Harchev for his quick response.
> >
> > It seems this is due to the fact that these characters are
> missing in fonts
> > AW uses (all fonts shipped with it have only glyphs for
> cp1252). It seems
> > GDI is choosing "best fit" replacements.
>
> No, this isn't it. I'm using a full Unicode truetype font
> (Code2000, by James Kass) which definitely has all the glyphs.
> Also, the replacements I see are not the typical "unknown
> character" glyphs that truetype fonts offer. They are the
> "best-fit" replacement characters that WideCharToMultibyte uses
> rather randomly (e.g. a Latin a for a Greek alpha, or a German
> sharp-s for a Greek beta.)
Are you really sure that AW uses that font? Is the proper font name is stored
in .abw file? Can you paste the same characters from the Word to AW without
loss of font information? -- I had a strong impression that AW can use only
fonts it ships. But I may be totally wrong.
I used AW for windows for no more than 3 minutes in my life, and I don't use
windows at all, so I'm not the right person to talk about hacking AW for
windows with.
> If Abiword makes GDI transform a Unicode string to Ansi just
> because its *own* fonts happen to be Ansi-only, it's doing a poor
> job.
>
> >
> > Please try 0.7.12 - it won't loose any character as I remember
> (I've coded
> > it).
>
> I'm afraid this *is* 0.7.12, downloaded yesterday. The sample
> file I used was test/wp/Unicode1.abw, which came with the 0.7.11
> download. I hate to say it, but most of the unicode characters
> were dropped.
> By the way, if you coded them not be dropped, then how? There
> must be a lot of Unicode characters for which standard LaTex
> equivalents simply don't exist?
On my unix box, the following chars got exported to latex equivs:
cyrillic (because I used russian locale) and greek ones and most card suites.
Conversion of unicode character to latex equivalent is performed by
a clone of the wv/text.c:wvConvertUnicodeToLaTeX
(I will make AW directly use a modified version of that function soon, so
there won't be a need to cut-n-paste that function, so if you wish to
contribute, modify only wvConvertUnicodeToLaTeX).
It would be trivial to add support for latin characters with unusual accented
marks (like encircled chars) and similar unicode symbols that can be
easily created using TeX constructs for combining symbols. Also I don't think
anybody will object to adding Omega (or what's the correct name for that
project) exporter - it will be trivial to add - see how RTF/RTF for old apps
exporter are implemented.
> Best regards,
>
> Lukas
>
Best regards,
-Vlad
This archive was generated by hypermail 2b25 : Thu Dec 21 2000 - 15:21:42 CST