[XeTeX] help with combining text/math

Will Robertson will at guerilla.net.au
Thu Feb 10 13:14:45 CET 2005

On 10 Feb 2005, at 1:47 PM, Scott Murman wrote:

> every few months i find a spare 1/2-hour to move further in the xetex 
> learning curve.  i've been able to get many of my files to process, 
> though i'm still missing some functionality.  i'm still trying to 
> understand the choices i have for mixing math & text.

Not many, unfortunately. As far as I'm aware, you can use CM, Euler, 
Lucida and possibly the tx/pxfonts (the fonts are there in OpenType but 
I haven't heard or investigate if they work).

> using fontspec to change the text font also changes some of the math 
> characters.   is it possible to disable this? at this point i'd be 
> happy to limit myself to the regular cm math fonts, and a light-stroke 
> text font.

I thought that fontspec only changes characters that occur in \mathrm 
and similar situations -- this means for things like \sin and \cos as 
well (*). If this is what you mean, then simply setting \setromanfont 
-after- the \begin{document} should have the desired effect of using 
CM. This is simply because math alphabets in LaTeX can only be defined 
before the beginning of the document.

(* That's what I thought, but now I see I'm wrong. How embarrassing. 
Actually the numbers and brackets and some things from the new font are 
also used in math. This is bad. I really really have to work out maths 
stuff better.)

Otherwise, post an example of what you are getting and an explanation 
of what you'd prefer, and I'll look into it!

> i've seen the xeuler package, and while i had trouble getting it 
> working, i'm not depressed about this since the euler fonts are 
> unattractive IMO.

The xeuler package only fixes the characters in the Euler Script fonts, 
which includes a few miscellaneous symbols. So depending on your needs 
\usepackage{euler} might be all you require. (Just in case you change 
your mind about their attractiveness :) I find them quite nice, but 
they are certainly very unusual.)

I hope that in the next XeTeX upgrade, the installer will take care of 
Euler Script automatically (not using the xeuler stuff, but rather 
using bsrx.map as I once posted). Of course, this assumes that the 
problem will still exist in the next version of XeTeX :)

Hope that helps,


More information about the XeTeX mailing list