# \mathsfup and \symsfup for variables

Unicode-math documentation v. 0.8o says in §3.1:

A boldface variable name should be printed using the math font, whereas a boldface operator name should be printed using the text font.

Now, how about typesetting multi-letter variables and multi-letter sorts (in other terminology, types) in multi-sorted logics? Which font should they be set up with, say, in the following example:

"For a variable identifier var : Int and a function identifier fun : Int → Bool the term fun(var) has sort Bool."

?

Let us continue:

The five new symbol font commands that behave in this way are: \symup, \symit, \symbf,\symsf , and \symit. These commands switch to single-letter mathematical symbols (generally within the same OpenType font). The legacy \math.. commands switch to text fonts that are set up to be- have correctly in mathematics, and should be used for multi-letter identifiers.

Ok, but what is, say, \mathsfup supposed to do? Notice that \mathsfup ∉ {\symup, \symit, \symbf,\symsf , and \symit} and \mathsfup is not a legacy command. Table 1 says \mathsfup is the same as \symsfup but what I am missing is the font used for \symsfup or \mathsfup.

Further, I cannot even find a proper definition of \symsfup:

$grep -air "symsfup" /usr/share/*tex*$


Where is it?

• Is \symfsup in the title of your query a typo? – Mico Apr 7 at 14:09
• For the record: I did not suggest the spelling \symfsup. To the contrary, I had asked you if there was a typo in the title of your posting. – Mico Apr 7 at 16:05

There is no explicit definition of \symsfup or other similar commands: they are implicitly defined as part of the set up.

Consider the following test file:

\documentclass{article}
\usepackage{unicode-math}

\pagestyle{empty}

\begin{document}

$\symsf{A}$ $\mathsf{A}$

$\symsfup{A}$ $\mathsfup{A}$

$\symbfsf{A}$ $\mathbfsf{A}$

$\symbfsfup{A}$ $\mathbfsfup{A}$

\showoutput
\end{document}


Isolating the relevant parts from the log file, we can see

....\mathon
....\TU/latinmodern-math.otf(1)/m/n/10 glyph#3126
....\mathoff
....\mathon
....\TU/lmss/m/n/10 glyph#27
....\mathoff

....\mathon
....\TU/latinmodern-math.otf(1)/m/n/10 glyph#3126
....\mathoff
....\mathon
....\TU/latinmodern-math.otf(1)/m/n/10 glyph#3126
....\mathoff

....\mathon
....\TU/latinmodern-math.otf(1)/m/n/10 glyph#3250
....\mathoff
....\mathon
....\TU/latinmodern-math.otf(1)/m/n/10 glyph#3250
....\mathoff

....\mathon
....\TU/latinmodern-math.otf(1)/m/n/10 glyph#3250
....\mathoff
....\mathon
....\TU/latinmodern-math.otf(1)/m/n/10 glyph#3250
....\mathoff


As you see, there is a difference only between \symsf and \mathsf, as written in the documentation.

In case you need to differentiate between \symbfsf and \mathbfsf, you can define the latter with

\setmathfontface⟨\mathbfsf⟩{<your text sans font>}[⟨the required features⟩]


so that multiletter strings in \mathbfsf are treated as if they were text, rather than juxtaposed symbols.

• @Just_A_Man Multiletter variables should use the text sans font (which will usually match the math sans font, of course). – egreg Apr 7 at 15:35
• @Just_A_Man Yes. Look at what you get from \symit{diff} and \mathit{diff}. – egreg Apr 7 at 15:42