Hi Hans and Mojca,


Thanks for the quick replies.

On 18 Jan 2006, at 15:22, Hans Hagen wrote:

> Mojca Miklavec wrote:
>
>> You don't need to put silly \uchar{}{} definitions in the text, I'm
>>
> indeed, uchar is not for this purpose and has special meaning in  
> cjk (mapping to special fonts) and xetex situations (mapping to  
> unicode)
>


Actually, I thought it was a good solution. Let me explain my current  
project.

I am typesetting a Festschrift including subjects on Indo-European  
(Sanskrit, Hittite, Old Greek, etc.), Slavic, Sino-Tibetan, and  
Japanese made up from 65 xml-files with decimal character entities.  
All the glyphs needed for these articles (save some silly ones like e  
with hungarian umlaut, m with subdot) are included in a font  
(Palatino Linostar) which was developed to accompany an etymological  
database (http://starling.rinet.ru/main.html) project devoted to  
proving or disproving the nostratic theory (all languages in the  
world originate from one source).   The font was developed over the  
years and is no longer fully unicode-compliant.

So my solution was to map the cmap hexadecimal values via the decimal  
entities to the \uchar command in a separate configuration file.

For example:

\defineXMLentity[257]{\uchar{1}{1}} % amacron % LATIN SMALL LETTER A  
WITH MACRON % 0x0101


I need about 2546 of them for all the special glyphs.

This solution works fine, except the fact that I am unable to typeset  
the glyphs in bold and italic.

Maybe this solution is a bit cumbersome. I am open for better solutions.

On 18 Jan 2006, at 13:10, Mojca Miklavec wrote:
> If the placement of the
> macron doesn't suit your font, you can redefine it and get rid of
> those \uchar-s.

What is the best way to redefine the accents (like macron, acute,  
etc.) to say a standard tex accent?

Is there a command to pick up certain utf-glyphs that aren't picked- 
up by context? See the  ƃ ǖ in my example file.

On 18 Jan 2006, at 13:10, Mojca Miklavec wrote:
> You can still use unicode or whatever input encoding
> (regime) you need (Latin4/6/7?). Send a complete minimal example to
> the list, including font definitions.

I have included a minimal testfile. I can send the pdf-output and  
zipfiles containing the encodingfiles and the linostarfont to you off- 
list since they are too big for the list.

I am using texexec version 5.4.3 with pdftex 3.141592-1.30.3-2.2  
(Web2C 7.5.5).


Sjoerd