How to make it so that newcommand takes the rest of my line as argument?

Question

I declared a new command to format list elements in LaTeX as follows:

\newcommand{\sitem}[1]{\item \makefirstuc{#1},}
\newcommand{\eitem}[1]{\item \makefirstuc{#1}.}

This works OK, but now I have to brace every list element as follows:

\begin{itemize}
    \sitem{First item}
    \eitem{second item}
\end{itemize}

How do I get it so that the following is possible?

\begin{itemize}
    \sitem First item
    \eitem second item
\end{itemize}

It is probably only possible if you leave a blank linie after each item, then you can grap up until the next \par token. But I'd say that \sitem{....} is easier to work with, especially when you later on want something slightly different. — daleif, Apr 19 '22 at 08:10
@daleif I itemize a lot in my documents, and the itemize environment is already quite verbose. Adding the new commands in the way I have done would hinder my workflow, so I really would like to find a way to do it without the braces. How is \item itself implemented? — user32882, Apr 19 '22 at 08:13
Are you willing and able to work with a LuaLaTeX-based approach? — Mico, Apr 19 '22 at 08:18
@Mico There's a package on CTAN called \mfirstuc which loads it in. It basically capitalizes the first letter of the item. — user32882, Apr 19 '22 at 08:33
For the "how \item is defined" part there's macros - Where do I find out how a command/environment is defined? - TeX - LaTeX Stack Exchange but you need some TeX programming skill to read the definition anyway. Basically it does something like "set indentation to X, check nesting level, print a '• / + / -' symbol" — user202729, Apr 19 '22 at 10:01
For a pure solution (apart from whole-file/regional textual substitution) maybe there's "setting catcode of end line char to something other than end of line catcode, grab argument until it, do the processing, set back". It's catcode-changing though, but there isn't another way — user202729, Apr 19 '22 at 10:02
@user202729 - You don't consider a LuaLaTeX-based solution to be "pure"? :-) — Mico, Apr 19 '22 at 11:18
See also macros - \def taking rest of the line as argument - TeX - LaTeX Stack Exchange — user202729, Feb 15 '23 at 16:29

score 5 · Accepted Answer · answered Apr 19 '22 at 08:45

5

Here's a LuaLaTeX-based solution. It sets up a Lua function, called process_items, which acts as a preprocessor: By the time TeX gets to process the input lines in question, it will not "see" \sitem first item but, instead, \item First item,. (Note the uppercasing of the first letter and the terminal comma.) The Lua function is activated and deactivated, respectively, by the LaTeX utility macros \ItemOn and \ItemOff.

% !TEX TS-program = lualatex
\documentclass{article} % or some other suitable document class
\usepackage{mfirstuc}   % for '\makefirstuc' macro
\usepackage{luacode}    % for 'luacode' environment
\begin{luacode}

function process_items ( s )
  s = string.gsub ( s , '\\sitem%s+(.+)' , '\\item\\makefirstuc{%1},' )
  s = string.gsub ( s , '\\eitem%s+(.+)' , '\\item\\makefirstuc{%1}.' )
  return s
end

\end{luacode}

\newcommand\ItemOn{\directlua{luatexbase.add_to_callback ( 
  "process_input_buffer", process_items , "ProcessItems" )}}
\newcommand\ItemOff{\directlua{luatexbase.remove_from_callback ( 
  "process_input_buffer", "ProcessItems" )}}

\begin{document}
\ItemOn %% assign the Lua function to the input processor callback

\begin{itemize}
    \sitem first item
    \sitem  second item
    \eitem   third item 
\end{itemize}
\end{document}

answered Apr 19 '22 at 08:45

Mico

506,678

1

It works. I don't fully understand how but the more I use it the more it'll sink in I guess. Why are you not using ItemOff in this example? – user32882 Apr 19 '22 at 13:19
1

@user32882 - I guess I created \ItemOff mainly as counterpart to the much more important \ItemOn macro. The code works because the Lua function is assigned to LuaTeX's process_input_buffer callback, meaning that it acts as a prepocessor and modifies matching patterns. By the time LaTeX gets to do its job, it never gets to "see" \sitem and \eitem. While \sitem and \eitem look like they might be LaTeX macros, that's not the case. Instead, their role is to serve as "markers" for pattern matching by Lua's powerful built-in string.gsub ("generalized substitution") function. – Mico Apr 19 '22 at 14:09
1

I'm afraid this breaks for multiple lines after the \sitem and \eitem. Any good ideas on how I can remedy that? – user32882 Apr 19 '22 at 14:41
2

@user32882 --- You asked for something that takes the rest of your line as an argument. If you want to allow multiple lines then TeX could only detect the end of the argument at the next instance of \sitem/\eitem or the \end{itemize}. Then you would have all sorts of fun and games with spaces between the argument and the full stop or comma. I suspect this is more trouble than it's worth. – Ian Thompson Apr 19 '22 at 15:02
@user32882 - In the title of your query, you specified "takes the rest of [the] line as [its] argument". That's exactly what my answer delivers. If that's not what you actually wanted, please clarify the condition(s) the input is supposed to satisfy. – Mico Apr 19 '22 at 15:05
@IanThompson this is worthwhile for me because I use itemize a lot – user32882 Apr 19 '22 at 15:05
1

@Mico I understand. I made another question: https://tex.stackexchange.com/questions/641312/lua-approach-for-itemize-fails-if-multiple-lines-are-given-after-item – user32882 Apr 19 '22 at 15:06

score 4 · Answer 2 · answered Apr 19 '22 at 15:01

You can borrow the \eoldef macro from OpTeX. Syntax:

\eoldef\macro #1{body of the macro, #1 can be used}

The \macro is defined with its parameter separated by the end of line.

In your LaTeX document it can look like this:

\documentclass{article}
\def\eoldef #1{\def #1{\begingroup \catcode\^^M=12 \eoldefA #1}% \expandafter\def\csname \string #1:M\endcsname} {\catcode^^M=12 %
\gdef\eoldefA #1#2^^M{\endgroup\csname \string #1:M\endcsname{#2}}%
}
\def\firstupper#1{\uppercase{#1}}
\eoldef\sitem#1{\item \firstupper #1,}
\eoldef\eitem#1{\item \firstupper #1.}
\begin{document}
\begin{itemize}
    \sitem First item
    \eitem second item
\end{itemize}
\end{document}

score 4 · Answer 3 · answered Apr 19 '22 at 22:25

If you don't plan to use nested lists in this case, you can implement the job in expl3.

Split the input at \item and massage the resulting sequence so that uppercasing of the first letter is performed. Then the whole lot is delivered adding ,\item in between items and a final period.

\documentclass{article}
\ExplSyntaxOn
\NewDocumentEnvironment{adjitemize}{+b}
 {
  \adjitemize_main:n { #1 }
 }
 {}
\seq_new:N \l__adjitemize_in_seq
\seq_new:N \l__adjitemize_out_seq
\cs_new_protected:Nn \adjitemize_main:n
 {
  % split the body at \item
  \seq_set_split:Nnn \l__adjitemize_in_seq { \item } { #1 }
  % discard the first item, which is empty
  \seq_pop_left:NN \l__adjitemize_in_seq \l_tmpa_seq
  % add the capitalization of the first letter
  \seq_set_map_x:NNn \l__adjitemize_out_seq \l__adjitemize_in_seq
   {
    \text_titlecase:n { ##1 }
   }
  % deliver the result
  \begin{itemize}
  \item \seq_use:Nn \l__adjitemize_out_seq { , \item } .
  \end{itemize}
 }
\ExplSyntaxOff
\begin{document}
\begin{adjitemize}
\item first item
      with a second line
\item second item
\item final item
\end{adjitemize}
\end{document}

I believe that using \item is much simpler than \sitem and \eitem, because you can move around the items without worrying about which one is last. The only thing to remember is to use adjitemize instead of itemize, but this is justified by the fact that you want to massage the items.

This is a valid answer to my other question. If you copy/paste it there I will accept it. — user32882, Apr 20 '22 at 06:42
Sure, I can delete it. But I'm keeping @Mico's answer for this question here. Your answer is also technically valid, and good reference for the future, but I still need to brush up my expl3 skills :) — user32882, Apr 20 '22 at 07:53

How to make it so that newcommand takes the rest of my line as argument?

3 Answers3