Combine numbers from custom commands into a range

Question

I’m defining datas in LaTeX with a custom command that stores both a label and a number:

\documentclass{article}
\usepackage{parskip}

\newcommand{\data}[3]{%
  \expandafter\newcommand\csname #1\endcsname[1][]{#2##1}%
  \expandafter\newcommand\csname #1Number\endcsname{#3}%
}

\newcommand{\getnumber}[1]{%
  \mbox{\textbf{(\csname #1Number\endcsname)}}%
}

\begin{document}

\data{tempA}{aaa}{1}
\data{tempB}{bbb}{2}
\data{tempC}{ccc}{3}
\data{tempD}{ddd}{4}
\data{tempE}{eee}{5}
\data{tempF}{fff}{6}

Data: \tempA \newline
Number: \getnumber{tempA}

\end{document}

For a single compound, \getnumber{tempA} outputs (1).

Now I want to create a command that takes a list of compound names and outputs a range summary of their numbers. For example:

\getnumberrange{tempA,tempB,tempC,tempF}  % should produce (1-3,6)

I’m not sure how to iterate over a list of command names, retrieve their numbers, and compress consecutive numbers into ranges. How can I implement this in LaTeX?

You could consider to store them into a l3clist, that is(is automatically rearrange the order) \getnumberrange{tempA,tempC,tempB,tempF} to a clist: 1,3,2,6, and be sorted 1,2,3,6, and use loops to judge whether the next item(clist[i+1]) of clist[i] is clist[i]+1, and replace them with -. But I was not sure what is the exact use of this command, it looks like duplicated wheels in cleveref or zref-clever or cite package.... — Explorer
– Explorer, Commented 2 days ago
is there a reason not to use latex's standard \label/\ref system here? as @Explorer said, then you could use e.g. zref-clever which can compress and format ranges etc. in addition to this, I think it would help if you could say a bit more about the end goal here. — cfr
– cfr, Commented 2 days ago
This quesion is the opposite/dual one of this: tex.stackexchange.com/q/494712/322482 — Explorer
– Explorer, Commented 2 days ago
Mandatory read for this: lambda-lists (documentation of lambda.sty) — jps
– jps, Commented yesterday

Explorer · Accepted Answer · 2025-12-19 06:17:17Z

Based on my comment, I give an expl3 attempt:

\documentclass{article}
\newcommand{\data}[3]{%
  \expandafter\newcommand\csname#1\endcsname[1][]{#2##1}%
  \expandafter\newcommand\csname#1Number\endcsname{#3}%
}
\ExplSyntaxOn
\cs_new:Npn \__taiwan_getnum:n #1
{
    \cs:w#1Number\cs_end:
}
\newcommand{\getnumber}[1]{%
%   \mbox{\textbf{(\__taiwan_getnum:n {#1})}}%
    \__taiwan_getnum:n {#1}%
}
\clist_new:N \l__taiwan_numrange_clist
\clist_new:N \l__taiwan_num_clist
\tl_new:N \l__taiwan_tmp_tl
\int_new:N \l__taiwan_tmp_int
\int_new:N \l__taiwan_start_int
\cs_new:Npn \__taiwan_getnumrange:n #1
{
    \group_begin:
    \clist_set:Nn \l__taiwan_numrange_clist {#1}
    \clist_clear:N \l__taiwan_num_clist
    \clist_map_inline:Nn \l__taiwan_numrange_clist
    {
        \clist_put_right:Nn \l__taiwan_num_clist {\__taiwan_getnum:n {##1}}%
    }
    \clist_sort:Nn \l__taiwan_num_clist
    {
        \int_compare:nNnTF { ##1 } > { ##2 }
        { \sort_return_swapped: }
        { \sort_return_same: }
    }
    \clist_pop:NN \l__taiwan_num_clist \l__taiwan_tmp_tl
    \int_set:Nn \l__taiwan_tmp_int { \l__taiwan_tmp_tl }
    \int_set:Nn \l__taiwan_start_int { \l__taiwan_tmp_tl }
    \clist_map_inline:Nn \l__taiwan_num_clist
    {
        \int_compare:nNnTF { ##1 } = { \int_eval:n { \l__taiwan_tmp_int + 1 } }
        {
            % consecutive
            \int_set:Nn \l__taiwan_tmp_int { ##1 }
        }
        {
            % not consecutive
            \int_compare:nNnTF { \l__taiwan_start_int } = { \l__taiwan_tmp_int }
            {
                % single number
                ~\int_use:N \l__taiwan_start_int ,
            }
            {
                % range
                ~ \int_use:N \l__taiwan_start_int -- \int_use:N \l__taiwan_tmp_int ,
            }
            \int_set:Nn \l__taiwan_start_int { ##1 }
            \int_set:Nn \l__taiwan_tmp_int { ##1 }
        }
    }
    \int_compare:nNnTF { \l__taiwan_start_int } = { \l__taiwan_tmp_int }
    {
        ~\int_use:N \l__taiwan_start_int
    }
    {
        ~ \int_use:N \l__taiwan_start_int -- \int_use:N \l__taiwan_tmp_int
    }
    \group_end:
}
\newcommand{\getnumberrange}[1]{%
    \__taiwan_getnumrange:n {#1}
}
\ExplSyntaxOff

\begin{document}

\data{tempA}{aaa}{1}
\data{tempB}{bbb}{2}
\data{tempC}{ccc}{3}
\data{tempD}{ddd}{4}
\data{tempE}{eee}{5}
\data{tempF}{fff}{6}

Data: \tempA 

Number: \getnumber{tempA} and \getnumber{tempD} 

Number Range: \getnumberrange{tempA,tempE,tempF,tempC}

Number Range: \getnumberrange{tempA,tempC,tempE}

Number Range: \getnumberrange{tempD,tempC}

Number Range: \getnumberrange{tempA,tempE,tempF,tempC,tempB,tempD}

Number Range: \getnumberrange{tempA,tempE,tempF,tempB,tempD}

\end{document}

Noted that I don't think I did a good job, as cfr also mentioned at here, there are similar achievement in zref-clever or cite to privide similar option named sort&compress, the above is probably reinvent-ing the wheels(NO!), so just regard it as my latex3 exercise. As for the () outside the outcome, you can add it on your own.

egreg · Accepted Answer · 2025-12-19 23:23:56Z

Here's my version with a few comments on the code.

I always avoid defining commands: it's better to have a separate namespace for such jobs.

\documentclass{article}

\ExplSyntaxOn

\NewDocumentCommand{\setdata}{mmm}
 {
  \prop_gput:Nnn \g_taiwan_data_prop {#1@name} {#2}
  \prop_gput:Nnn \g_taiwan_data_prop {#1@number} {#3}
 }
\NewExpandableDocumentCommand{\getname}{m}
 {
  \taiwan_data_name:n {#1}
 }
\NewDocumentCommand{\getdata}{m}
 {
  \taiwan_data_get:n {#1}
 }

\prop_new:N \g_taiwan_data_prop
\seq_new:N \l__taiwan_data_items_seq
\seq_new:N \l__taiwan_data_ranges_seq
\seq_new:N \l__taiwan_data_ranges_final_seq
\tl_new:N \l__taiwan_data_collect_tl
\int_new:N \l__taiwan_data_previous_int

\cs_new:Nn \taiwan_data_name:n
 {
  \prop_item:Nn \g_taiwan_data_prop {#1@name}
 }

\cs_new_protected:Nn \taiwan_data_get:n
 {
  (
  \int_compare:nTF { \clist_count:n {#1} = 1 }
   {% if just one item, produce it
    \prop_item:Nn \g_taiwan_data_prop {#1@number}
   }
   {% otherwise, sort and print
    \__taiwan_data_sort:n {#1}
   }
  )
 }

\cs_new_protected:Nn \__taiwan_data_sort:n
 {
  % first phase: collect the numbers
  \seq_clear:N \l__taiwan_data_items_seq
  \clist_map_inline:nn {#1}
   {
    \seq_put_right:Ne \l__taiwan_data_items_seq { \prop_item:Nn \g_taiwan_data_prop {##1@number} }
   }
  % second phase: sort the sequence
  \seq_sort:Nn \l__taiwan_data_items_seq
   {
    \int_compare:nNnTF {##1}>{##2}
     { \sort_return_swapped: }
     { \sort_return_same: }
   }
  % third phase: print
  % 1. detach the first item
  \seq_pop_left:NN \l__taiwan_data_items_seq \l__taiwan_data_collect_tl
  \int_set:Nn \l__taiwan_data_previous_int { \l__taiwan_data_collect_tl }
  % 2. start the mapping
  \seq_map_inline:Nn \l__taiwan_data_items_seq
   {
    \int_compare:nTF { ##1 == \l__taiwan_data_previous_int + 1 }
     {% we're in a range
      \tl_put_right:Nn \l__taiwan_data_collect_tl { , }
     }
     {% not in a range
      \tl_put_right:Nn \l__taiwan_data_collect_tl { // }
     }
    \tl_put_right:Nn \l__taiwan_data_collect_tl {##1}
    \int_set:Nn \l__taiwan_data_previous_int {##1}
   }
  % 3. split at //
  \seq_set_split:NnV \l__taiwan_data_ranges_seq { // } \l__taiwan_data_collect_tl
  % 4. massage the possible ranges
  \seq_clear:N \l__taiwan_data_ranges_final_seq
  \seq_map_inline:Nn \l__taiwan_data_ranges_seq
   {
    \int_compare:nTF { \clist_count:n {##1} > 1 }
     {% we have a range
      \seq_put_right:Ne \l__taiwan_data_ranges_final_seq
       { \clist_item:nn {##1} {1} -- \clist_item:nn {##1} {-1} }
     }
     {% no range
      \seq_put_right:Nn \l__taiwan_data_ranges_final_seq {##1}
     }
   }
  % 5. print
  \seq_use:Nn \l__taiwan_data_ranges_final_seq {,~}
 } 

\ExplSyntaxOff

\begin{document}

\setdata{tempA}{aaa}{1}
\setdata{tempB}{bbb}{2}
\setdata{tempC}{ccc}{3}
\setdata{tempD}{ddd}{4}
\setdata{tempE}{eee}{5}
\setdata{tempF}{fff}{6}


Data: \getname{tempA} -- Number: \getdata{tempA}

\getdata{tempA,tempB,tempC,tempF}

\getdata{tempA,tempB,tempC,tempE,tempF}

\getdata{tempB,tempF,tempE,tempA,tempC} % same as before

\end{document}

jps · Accepted Answer · 2025-12-19 22:06:12Z

Just because it's possible, the following defines a fully expandable version that retrieves the numeric values, then sorts them, and finally turns them into a comma separated list with ranges detected and replaced by <start>--<end>. If <end> is only bigger than <start> by 1 it'll not be considered a range.

\documentclass{article}

\ExplSyntaxOn
\NewDocumentCommand \data { m m m }
  {
    \ExpandArgs{c} \NewDocumentCommand {#1} { O{} } { #2 ##1 }
    \cs_new:cpx { taiwan~number-\tl_to_str:n {#1} } { \int_eval:n {#3} }
  }
\NewExpandableDocumentCommand \getnumber { m }
  { ( \exp_after:wN \use:n \__taiwan_getnumber:n {#1} ) }
\NewExpandableDocumentCommand \getnumberrange { m }
  { ( \taiwan_getnumber_range:n {#1} ) }

% Turn a clist of data labels into a simplified clist of numbers possibly
% containing ranges.
\cs_new:Npn \taiwan_getnumber_range:n #1
  {
    % #1: clist of data labels
    \exp_last_unbraced:Ne \use_none:n % remove leading ,
      {
        \taiwan_rangify:e
          {
            \tl_sort:eN
              { \clist_map_function:nN {#1} \__taiwan_getnumber:n }
              \__taiwan_if_smaller_than:nnTF
          }
      }
  }
\cs_generate_variant:Nn \tl_sort:nN { e }
\cs_new:Npn \__taiwan_if_smaller_than:nnTF #1 { \int_compare:nNnTF {#1} < }
% Turn a single data label into a number inside of braces (so into an item of a
% tl).
\cs_new:Npn \__taiwan_getnumber:n #1
  {
    {
      \cs_if_exist_use:cF { taiwan~number-\tl_to_str:n {#1} }
        { \msg_expandable_error:nnn { taiwan } { unknown-number } {#1} 0 }
    }
  }

% Simplify a tl of numbers into a clist possibly containing ranges.
\cs_new:Npn \taiwan_rangify:n #1
  {
    % #1: tl with each item being a number
    % we use \q_nil to denote the end of the list and \q_stop as a marker to
    % gobble the remainder of the current loop iteration inside
    % \__taiwan_rangify:w.
    \__taiwan_rangify:w #1 \q_nil \q_stop
  }
\cs_generate_variant:Nn \taiwan_rangify:n { e }
\cs_new:Npn \__taiwan_rangify:w #1
  {
    % #1: either \q_nil if the tl provided to \taiwan_rangify:n was blank or the
    %     first number
    % if #1 is \q_nil we need to leave a single token for the \use_none:n in
    % \taiwan_getnumber_range:n.
    \use_none_delimit_by_q_nil:w #1 , \use_none_delimit_by_q_stop:w \q_nil
    \__taiwan_rangify_aux:nw {#1}
  }
\cs_new:Npn \__taiwan_rangify_aux:nw #1 #2
  {
    % #1: a number
    % #2: either \q_nil or the next number from the list >= #1
    \use_none_delimit_by_q_nil:w #2 ,#1 \use_none_delimit_by_q_stop:w \q_nil
    \int_case:nnF {#2}
      {
        { #1 }     { \__taiwan_rangify_aux:nw {#1} }
        { #1 + 1 } { \__taiwan_rangify_aux:Nnnw \use_i:nn {#1} {#2} }
      }
      { , #1 \__taiwan_rangify_aux:nw {#2} }
  }
\cs_new:Npn \__taiwan_rangify_aux:Nnnw #1#2#3 #4
  {
    % #1: \use_i:nn if #2 + 1 = #3, \use_ii:nn if #2 + 1 < #3
    % #2: a number
    % #3: a number bigger than #2
    % #4: either \q_nil or the next number from the list >= #3
    \use_none_delimit_by_q_nil:w #4
      \__taiwan_rangify_range:Nnn #1 {#2} {#3}
      \use_none_delimit_by_q_stop:w
      \q_nil
    \int_case:nnF {#4}
      {
        { #3 }     { \__taiwan_rangify_aux:Nnnw #1 {#2} {#3} }
        { #3 + 1 } { \__taiwan_rangify_aux:Nnnw \use_ii:nn {#2} {#4} }
      }
      {
        \__taiwan_rangify_range:Nnn #1 {#2} {#3}
        \__taiwan_rangify_aux:nw {#4}
      }
  }
% output two numbers either as a range (if #2 + 1 < #3) or as a list
\cs_new:Npn \__taiwan_rangify_range:Nnn #1#2#3
  {
    % #1: \use_i:nn if #2 + 1 = #3, \use_ii:nn if #2 + 1 < #3
    % #2: a number
    % #3: a number bigger than #2
    , #2 #1 { , } { -- } #3
  }
\msg_new:nnn { taiwan } { unknown-number }
  { Unknown~ number~ #1.~ Using~ 0~ instead. }
\ExplSyntaxOff

\begin{document}
\data{tempA}{aaa}{1}
\data{tempB}{bbb}{2}
\data{tempC}{ccc}{3}
\data{tempD}{ddd}{4}
\data{tempE}{eee}{5}
\data{tempF}{fff}{6}

\tempA: \getnumber{tempA}

\tempA: \getnumberrange{tempA}

\getnumberrange{tempA,tempF,tempE,tempB,tempC,tempA,tempA}

% proof things are expandable
\edef\foo{\getnumberrange{tempA,tempF,tempE,tempB,tempC,tempA,tempA}}
\texttt{\meaning\foo}
\end{document}

Stack Exchange Network

Combine numbers from custom commands into a range

3 Answers 3

You must log in to answer this question.

Linked

Hot Network Questions

Combine numbers from custom commands into a range

3 Answers 3

You must log in to answer this question.

Linked

Related

Hot Network Questions