Lua Latex：如何在内联模式下打破集合中一长行元素？

Question

我有一个很长的 2-uple 元素列表，例如$A = \{ (1,2), (1,3), (1,5), (2,1), (2,2), (2,3), (1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3), \}$使用内联数学模式。LuaLaTeX 不会分成多行。我如何才能强制它在不在元组内的逗号处分行，否则集合元素会超出页面？

提出的解决方案在内联数学模式中允许在“，”处换行吗？没用。似乎破坏了多语性。

Answer 1

这里有一种可能性，即使用名为 的命令，\mySpecialCommas仅在使用该命令时，逗号才会在数学模式下处于活动状态。当\mySpecialCommas在数学公式中使用时，逗号后允许换行，当且仅当逗号后跟有空格标记（通常可以从行尾获取）。因此，请确保元组中的逗号后没有空格。请注意示例中的最后一个逗号，除非您同意在行首出现右括号。这应该适用于任何引擎。

\documentclass{article}
\usepackage{lipsum}

\makeatletter
\newcommand*{\my@specialComma}{%
  \futurelet\next\my@specialComma@next
}

\newcommand*{\my@specialComma@next}{%
  \mathchar "613B             % the normal mathematical comma
  \ifx\next\@sptoken
    \penalty 0                % higher values make a line break less desirable
    \mskip 0mu plus 2mu\relax % add stretchability for better line breaks
  \fi
}

{\catcode`\,=\active
 \gdef,{\my@specialComma}%
}

\newcommand*{\mySpecialCommas}{\mathcode`\,="8000 } %space intended
\makeatother

\begin{document}

\lipsum*[1][1-3]
This is a long equation: $\mySpecialCommas
A = \{ (1,2), (1,3), (1,5), (2,1), (2,2), (2,3), (1,2), (1,3), (1,5), (2,1),
(2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5),
(2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3) \}$.

\medskip
This is a long equation: $\mySpecialCommas
A = \{ (1,2), (1,3), (1,5), (2,1), (2,2), (2,3), (1,2), (1,3), (1,5), (2,1),
(2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5),
(2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3),\}$.

\medskip
Normal text can have commas, no problem.

\end{document}

Answer 2

既然您提到您使用 LuaLaTeX，那么这里有一个解决方案，其工作原理是设置（a）一个 Lua 函数，\allowbreak如果逗号跟在)字符后面，则插入指令，以及（b）一个名为 的 LaTeX 前端宏\AllowBreaks。

上图的垂直线表示文本块的边缘。

% !TEX TS-program = lualatex
\documentclass{article}
\usepackage{showframe}
\usepackage{luacode} % for '\luaexec' and '\luastring' macros
\luaexec{
function AllowBreaks ( s )
  tex.sprint ( ( s:gsub ( "\%)\%s-," , "),\\hspace{1sp}\\allowbreak" ) )  )
end
}
\newcommand\AllowBreaks[1]{\luaexec{AllowBreaks(\luastring{#1})}}

\newcommand\LongString{test test test $A = \{ (1,2), (1,3), (1,5), (2,1), (2,2), (2,3), (1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3), (1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3) \}$ test test test test}

\begin{document}
\LongString

\bigskip
\AllowBreaks{\LongString}
\end{document}

Answer 3

这个答案确实不是require lualatex，而是依靠 tokcycle包来遍历环境内容的标记流，并\allowbreak在括号组之外的逗号后添加。此外，为了模拟\sloppy输入的某些内容，它会添加一些带有空格的额外胶水以允许拉伸/压缩（您可以调整数量）。

\documentclass{article}
\usepackage[showframe,pass]{geometry}
\usepackage{tokcycle}
\Characterdirective{\ifx(#1\def\commaaction{}\else
  \ifx)#1\def\commaaction{\allowbreak}\fi\fi
  \addcytoks{#1}\ifx,#1\addcytoks[1]{\commaaction}\fi}
\Spacedirective{\addcytoks{\hspace{0pt plus 2pt minus 1pt}#1}}
\begin{document}
\tokencyclexpress
$A = \{ (1,2), (1,3), (1,5), (2,1), (2,2), (2,3), (1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3),(1,2), (1,3), (1,5), (2,1), (2,2), (2,3), \}$
\endtokencyclexpress
\end{document}