Funny behavior with precedences

[kohlhase]

There is still a problem with bracketing, ... Here is a MWE:

\documentclass{stex}
\begin{document}
\begin{smodule}{foo}
  \symdef{limpl}[args=ai,op=\Rightarrow,prec=500;499x499,assoc=binr,name=implication]
  {#1\comp\Rightarrow #2}{##1\comp{\Rightarrow}##2}
  \def\bC{\mathbf{C}}
%  \vardef{bC}{\comp C}
    $\limpl{\limpl{\bC,\limpl\bC\bC}\bC,\limpl{\bC,\bC}\bC,\bC}\bC$
    \end{smodule}
\end{document}

When I use the \def\bC{\mathbf{C}}, then I get

(note the missing brackets) and when I use \vardef{bC}{\comp C} I get

Is this really intended?

[Jazzpirate]

this is indeed weird - since either way the brackets are around an \limpl\bC\bC, whether to insert them should be governed by \limpl rather than \bC, so one would think that the definition of \bC doesn't matter...

[Jazzpirate]

I will recheck the bracketing-algorithm for associative arguments, I'm assuming it treats the whole argument sequence as one (accumulated) argument. Not entirely sure why that would cause the problem, but that seems like a likely cause to me. If that's the problem, I'll fix it for sTeX 3.4

[kohlhase]

That is soon enough, I know how to get the behavior I want.