A strange bug related to equation environment

[Clark-Young]

When I try add below to my .tex file and compile it,

To understand how to calculate the exchange correlation density, let's clarify some definitions. The exchange correlation energy density is defined as \begin{equation}
  E[\rho] = \int \rho(\bfit{r}) \varepsilon[\rho, \nabla \rho, \nabla^2 \rho, \cdots] (\bfit{r})\mathrm{d} \bfit{r}
\end{equation}

I got an error:
Bad register code (32768).
__regex_query_set_aux:nN ...__regex_curr_pos_int

But when I try to get an mwe like

\documentclass[11pt, paperstyle=light yellow, color entry,
  title in boldface, title in sffamily, use style = classical]{jwjournal}
% paperstyle = ... 用于调节纸张颜色，可选项包括：lightyellow、yellow、parchment、green、lightgray、gray、nord、dark
% color entry 可以为每个条目的标题增添更多色彩
% scroll 可以开启卷轴模式，生成类似于长截图的单页 pdf
\usepackage{unicode-math}
\usepackage{amsmath}
\UseLanguage { English } % 设置文档语言为英语
\newcommand{\bfit}[1]{\symbfit{#1}}
\begin{document}


2025-07-08 Sunny | Anywhere

2025-08-28 Cloudy | French Family Science Center

[More about the Exchange Correlation Density]

To understand how to calculate the exchange correlation density, let's clarify some definitions. The exchange correlation energy density is defined as \begin{equation}
  E[\rho] = \int \rho(\symbfit{r}) \varepsilon[\rho, \nabla, \nabla^2, \cdots] (\bfit{r})\mathrm{d} \bfit{r}
\end{equation}

\end{document}

I could complile it successfully. No bugs arised.

and if I change the equation environment to $$ $$, no bugs arised.

If I add subscripts like,

$$
E_{xc}[\rho] = \int \rho(\symbfit{r}) \varepsilon_{xc}[\rho, \nabla \rho, \nabla^2 \rho, \cdots] (\symbfit{r})\mathrm{d} \symbfit{r}
$$

the bug arrised again.

I'm not sure what happened here. If you need more information, please let me know.

[Jinwen-XU]

There seems nothing wrong with your code. The error message appears to indicate that the current content is too long for the default regex mechanism to parse, but without a complete document to reproduce the error I am not 100% sure of this. The error is probably triggered by the extra code trying to skip the parsing process of math content inside the \[...\], and with $$...$$ this part of code would be ignored (which usually causes trouble), thus the fact that you may still compile the document is purely luck since your current formula does not contain elements that might break the regex scanning (in other words, try avoiding $$...$$ in future when using LaTeX, since by design \[...\] should always be better than $$...$$ in LaTeX document).

If this is indeed the reason, then I might not be able to fix this (I think the only way to make it work for long documents is to rewrite the code with Lua). Just try to separate the code into subfiles and input them via :: {⟨file name⟩}. Let me know if this works;)

[Clark-Young]

I do have a long document. And your suggestion works! Thank you.