Bug in mathic option, use of inproperly initialised counter

[daleif]

A user sent me this MWE

\documentclass{article}
\usepackage{mathtools}
\mathtoolsset{mathic}
\usepackage[left=6.6cm, right=6.6cm]{geometry}
\begin{document}
\textit{A connected graph~\(G\) on \(n\) vertices has no less than \(n-1\) edges.}\par
\textit{A connected graph \(G\) on \(n\) vertices has no less than \(n-1\) edges.}
\end{document}

I'll explain (credit to David in chat) in a response

[daleif]

The problem is the following code behind the mathic option

\def\MT_mathic_true: {
  \MH_if_boolean:nF {math_italic_corr}{
    \MH_set_boolean_T:n {math_italic_corr}
    \MH_if_boolean:nTF {robustify}{
      \MH_let:NwN \MT_mathic_redeffer: \DeclareRobustCommand
    }{
      \MH_let:NwN \MT_mathic_redeffer: \renewcommand
    }
    \MH_let:NwN \MT_begin_inlinemath: \(
    %\renewcommand*\({
    \MT_mathic_redeffer:*\({
      \relax\ifmmode\@badmath\else
      \ifhmode
        \MH_if_dim:w \fontdimen\@ne\font>\z@
          \MH_if_dim:w \lastskip>\z@
            \skip@\lastskip\unskip
            \MH_if_num:w \lastpenalty>\z@
              \count@\lastpenalty\unpenalty
            \MH_fi:
            \@@italiccorr
            \MH_if_num:w \count@>\z@ %      <----- bug
              \penalty\count@
            \MH_fi:
            \hskip\skip@
          \MH_else:
            \@@italiccorr
          \MH_fi:
        \MH_fi:
      \fi
      $\fi
    }
  }
}

The penalty in the marked position is inserted just as long as \count@ is non-zero, but the code that we are reacting to never sets the scratch counter if that test fails.

For now my suggested fix is to simply use

   \MH_if_num:w \lastpenalty>\z@
              \count@\lastpenalty\unpenalty
   \MH_else:
              \count@\z@
   \MH_fi:

to ensure that the counter is properly initialised to work with the test after the italic correction.

David notes in chat, that given we assume etex, it will be better to do the test using \ifnum\lastnodetype=13 instead of the \ifnum\lastpenalty>\z@ test. I'll mark that as a comment in the DTX and leave the cleanup to when mathtools is converted into Expl3.