I have a XeTeX source which compiles "nicely" into PDF except for many messages of the following kind
Missing character: There is no in font [Asana-Math]/ICU:script=math;language=
DFLT;!
First of all, I would like to know is it possible to identify the line number with a character which is missing? My document is already big enough (~150 A4 PDF pages long) and contains a lot of math, so it is almost impossible to isolate such problems.
Secondly, I located one particular case when it happens and I'm unpleasantly surprised. The MWE is as follows
\documentclass{article}
\usepackage{unicode-math}
%\setmathfont{[Asana-Math]}
\begin{document}
$\mathbfup{Z}^{+}$
\end{document}
Font used to typeset math does not matter, at least on my TeX Live 2013 installation the result is the same with both Latin Modern Math (default) and Asana Math fonts.
And surprise is that if you remove curly brackets over +, the corresponding Missing character message disappears from the log file. I always thought that ^{+} is semantically the same as ^+, since you can omit curly brackets when subscript or superscript is just 1 symbol long.
But it feels like there is a difference.