[XeTeX] again russian, german, english...

[Jonathan Kew]

On 10 Feb 2006, at 7:31 am, Martin Henning wrote:

> Hello list,
>
> i'm really sorry to have to bug you again with this topic, but i  
> simply can't get it to work properly. I put together a test page,  
> which contains my current setup and illustrates my problems. It  
> should be processed with XeLaTeX.
>
> So far i have the following things: Somehow the language selection  
> seems to work, as i can see how to fonts change. The english  
> variant looks fine to me, at least it is more or less beautiful.  
> The Russian variant seems to be pretty messed up though. Huge font,  
> possibly bold, a lot more space between lines and as the \itemize  
> on the third page shows, very ugly lists. The German font looks ok,  
> though i loose all my Umlauts, which makes it kinda of useless to  
> use German :)) Finally i get warnings that something is not allowed  
> in PDFEncodedString - i guess that's why the last problem arises:  
> PDF-Menu doesn't show Cyrillic. I have seen it in other PDFs so it  
> seems to be possible, just don't know how! My paper is due on March  
> 1st, so i'm hoping it will be enough time for me to make it  
> beautiful :)

I don't have much time this morning, but will try to review this at  
the weekend. A couple observations up front, though (and maybe  
someone else can help with solutions.... these are LaTeX/Babel  
matters, where I'm really out of my depth).

1. The \documentclass{...} line is missing from your sample, so I ran  
it as "article", but a couple of the commands aren't defined there.  
So we may not be seeing exactly what you're seeing -- but the general  
problems will be the same, so it's not a big deal.

2. The German and English are not appearing in the fontspec font, but  
defaulting to CM. This is probably a case of Babel overriding the  
encoding/font choices that fontspec tries to make. And that's why the  
umlauts disappear--you end up trying to print Unicode umlauted  
characters using a CM font that doesn't support those codes. I recall  
we tricked Babel for Russian by defining \cyrillicencoding, but am  
not sure yet how to make it keep its hands off the Latin-script text  
as well.

3. To get "close" line spacing with Charis SIL, you need to add  
something like
	\setlength{\lineskiplimit}{-4pt}
to the preamble, as Charis SIL has very large ascent/descent values  
in the font (to allow space for tall/deep characters and stacked  
diacritics).

4. By the PDF-Menu, you mean bookmarks? I don't think you can get non- 
ASCII characters in there with XeTeX at the moment, because of  
limitations in the Perl script that xdv2pdf uses to add the  
bookmarks. In theory, someone with deep enough PDF knowledge (not  
me!) might be able to add Unicode support there, but I'm expecting to  
roll out a better solution sometime fairly soon....

JK