Dear list,

I am trying to typeset some multi-letter words/identifiers in mathematical formulas. Currently, doing things like {\it variable} or {\mathit variable} under math mode will cause the engine to shift all characters into the corresponding unicode mathematical alphanumeric block position. However, most OpenType fonts that have good support for unicode math are optimized for single-character usage. Therefore, the outcome of such method is non-ideal, as no kerning or ligatures will be enabled. Especially, this gives bad output when typesetting variable names with combinations like “ffi”.

Looking through some past discussions on the list, I am aware of the usage of \mathword and \mathtext. Using those command will keep all the characters “as is” and typeset them under text mode (as far as I can tell). But those commands switch to the main body fonts temporarily to accomplish this. Sometime when using a different math font than the main font (like Palatino with Latin Modern Math), I wish my variable names can also be in the math font (or at least very similar, such as Latin Modern vs. Latin Modern Math) to better merge with the formula.

The way I am currently doing is to introduce some “math text fonts” that is different from the main text font but close to the math font. And with custom font switchers like \mwrm and \mwit to help me define things like

\def\wordrm#1{\mbox{\mwrm #1}}
\def\wordit#1{\mbox{\mwit #1}}

Am I missing some builtin functionalities or easier ways to do this? If not, would it be a good idea to extend the functionality of \mathword so it lifts out the restriction of only using the main text font?

Thanks!

Shuxian Wang