Index for a multi-volume book

Question

I want to create an index for a multi-volume book. I tried this one: How to produce a combined index from multiple documents? There are some first results, but the sorting is wrong. Here is a short example:

First volume:

\documentclass{article}
\usepackage{makeidx}
\makeindex

\newcommand\docname[1]{#1}
\let\LATEXindex\index % save old definition to prevent recursion
\renewcommand\index[1]{\LATEXindex{#1|docname{I}}}

\begin{document}

\index{word1}word1
\index{word2}word2
\index{word3}word3
\index{word4}
\newpage
\setcounter{page}{3}
\index{word1}word1
\newpage
\setcounter{page}{4}
\index{word1}word1
\index{word2}word2
\newpage
\setcounter{page}{5}
\index{word1}word1
\index{word2}word2

\printindex
\end{document}

second volume:

\documentclass{article}
\usepackage{makeidx}
\makeindex

\newcommand\docname[1]{#1}
\let\LATEXindex\index % save old definition to prevent recursion
\renewcommand\index[1]{\LATEXindex{#1|docname{II}}}

\begin{document}

\index{word1}word1
\index{word3}word3
\index{word4}word4
\newpage
\setcounter{page}{3}
\index{word1}word1
\index{word2}word2
\index{word3}word3
\newpage
\setcounter{page}{4}
\index{word1}word1
\index{word3}word3
\newpage
\setcounter{page}{5}
\index{word2}word2
\index{word4}word4

\printindex
\end{document}

index volume:

\documentclass{article}
\usepackage{makeidx}

\newcommand\docname[2]{#1: #2}

\begin{document}

\printindex
\end{document}

As written in the above link, the two .idx documents are combined to one, I do it by:

copy /b *.idx volume_ind.idx

After that, do makeindex and latex on volume_ind.idx. It works, but the two volumes are intermixed and the sorting is wrong. It looks like this:

Does anybody have an idea how to get the ordering of the index entries correct? Best would be something like this:

word1 I: 1, 3-4, II: 1, 3-4

Thank you a lot!!

Answer 1

I found an answer myself. Now only one document is needed and a perl script.

\documentclass{article}
\usepackage{makeidx}
\usepackage{xpatch}
\makeindex

\makeatletter
\patchcmd{\@wrindex}{\thepage}{\volumenr-\thepage}{}{}
\makeatother

\begin{document}
\newcommand{\volumenr}{I}

\index{word1}word1
\index{word2}word2
\index{word3}word3
\index{word4}
\newpage
\setcounter{page}{3}
\index{word1}word1
\newpage
\index{word1}word1
\index{word2}word2
\newpage
\index{word1}word1
\index{word2}word2

\renewcommand{\volumenr}{II}

\index{word1}word1
\index{word3}word3
\index{word4}word4
\newpage
\setcounter{page}{3}
\index{word1}word1
\index{word2}word2
\index{word3}word3
\newpage
\index{word1}word1
\index{word3}word3
\newpage
\index{word2}word2
\index{word4}word4

\printindex
\end{document}

The perl script eliminates the redundant roman numbers in the *.ind file, so I keep only one "I-" or "II-" in a line:

#!/usr/bin/perl
use strict;

my $src = shift @ARGV;
my $des = shift @ARGV;

open(SRC, '<', $src) or die $!;
open(DES,'>',$des) or die $!;

while (my $row = <SRC>) {
  $row =~ s/II-/II: /;          # search first "II-" in each row, replace with "II: "
  $row =~ s/(?<!%)\II-//g;      # search all "II-" in each row, delete

  $row =~ s/I-/I: /;            # search first "I-" in each row, replace with "I: "
  $row =~ s/(?<!%)\I-//g;       # search all "I-" in each row, delete
  print DES "$row";
}

close(SRC);
close(DES);

I call the perl script after makeindex and before the second latex call.

My index now looks like this:

Answer 2

To do that I recommend glossaries-extra with bib2gls. For example, if you have this three documents:

MyDoc1.tex

\documentclass{article}

\usepackage[record,index,postdot]{glossaries-extra}
\usepackage{glossary-bookindex}
\setglossarystyle{bookindex}

%number of columns
\renewcommand{\glsxtrbookindexcols}{2}

%names for glossaries, there can be more than one
\newglossary{symbols}{lds}{lsi}{Symbols List}
\newglossary{index}{ing}{ige}{General Index}

%setting external documents and prefixes
\newcommand{\setfileprefix}[2]{\csdef{prefix@#1}{#2}}
\newcommand{\usefileprefix}[1]{\csuse{prefix@#1}}

\setfileprefix{MyDoc2.pdf}{II-}
\setfileprefix{MyDoc3.pdf}{III-}

\renewcommand*\glsxtrdisplaysupploc[5]{%
  \setentrycounter[#1]{#2}%
  \usefileprefix{#4}\glsxtrmultisupplocation{#5}{#4}{#3}%
}

%style of symbols list
\newglossarystyle{listsimb}{%
\glossarystyle{super}%
\renewenvironment{theglossary}%
    {\tablehead{}\tabletail{}%
     \begin{supertabular}{@{}lp{20cm}}}%
    {\end{supertabular}}%
\renewcommand{\glsgroupskip}{}%
\renewcommand*{\glossaryentryfield}[5]{%
\glsentryitem{##1}\glstarget{##1}{##2} & ##3\glspostdescription\space ##5\\[2pt]}%
}


%font
\renewcommand*{\glsnamefont}[1]{\textmd{#1}}

%loading supplemental files
\GlsXtrLoadResources[
 src=index,
 type=index,
 supplemental-locations={MyDoc2, MyDoc3},% you can have more than one!
 supplemental-category={supplemental},
 supplemental-selection={all}
]

%and you can define all kinds of glossaries
\GlsXtrLoadResources[
 src=symbols,
 type=symbols,
 supplemental-locations={MyDoc2, MyDoc3},
 supplemental-category={supplemental},
 supplemental-selection={all}
]

\begin{document}
A test \gls{ind1} and another time with different output \glsuseri{ind1}. Finally a constant \gls{symb1}.

\printunsrtglossary[type=index, style=bookindex]
\printunsrtglossary[type=symbols, style=listsimb]
\end{document}

MyDoc2.tex

\documentclass{article}

\usepackage[record,index,postdot]{glossaries-extra}
\usepackage{glossary-bookindex}
\setglossarystyle{bookindex}

%number of columns
\renewcommand{\glsxtrbookindexcols}{2}

%names for glossaries, there can be more than one
\newglossary{symbols}{lds}{lsi}{Symbols List}
\newglossary{index}{ing}{ige}{General Index}

%setting external documents and prefixes
\newcommand{\setfileprefix}[2]{\csdef{prefix@#1}{#2}}
\newcommand{\usefileprefix}[1]{\csuse{prefix@#1}}

\setfileprefix{MyDoc1.pdf}{I-}
\setfileprefix{MyDoc3.pdf}{III-}

\renewcommand*\glsxtrdisplaysupploc[5]{%
  \setentrycounter[#1]{#2}%
  \usefileprefix{#4}\glsxtrmultisupplocation{#5}{#4}{#3}%
}

 %style of symbols list
\newglossarystyle{listsimb}{%
\glossarystyle{super}%
\renewenvironment{theglossary}%
    {\tablehead{}\tabletail{}%
     \begin{supertabular}{@{}lp{20cm}}}%
    {\end{supertabular}}%
\renewcommand{\glsgroupskip}{}%
\renewcommand*{\glossaryentryfield}[5]{%
\glsentryitem{##1}\glstarget{##1}{##2} & ##3\glspostdescription\space ##5\\[2pt]}%
}

%font
\renewcommand*{\glsnamefont}[1]{\textmd{#1}}

%loading supplemental files
\GlsXtrLoadResources[
 src=index,
 type=index,
 supplemental-locations={MyDoc1, MyDoc3},% you can have more than one!
 supplemental-category={supplemental},
 supplemental-selection={all}
]

%and you can define all kinds of glossaries
\GlsXtrLoadResources[
 src=symbols,
 type=symbols,
 supplemental-locations={MyDoc1, MyDoc3},
 supplemental-category={supplemental},
 supplemental-selection={all}
]

\begin{document}
A test in this document \gls{ind2} and a constant \gls{symb1}.

\printunsrtglossary[type=index, style=bookindex]
\printunsrtglossary[type=symbols, style=listsimb]
\end{document}

and MyDoc3.tex

\documentclass{article}

\usepackage[record,index,postdot]{glossaries-extra}
\usepackage{glossary-bookindex}
\setglossarystyle{bookindex}

%number of columns
\renewcommand{\glsxtrbookindexcols}{2}

%names for glossaries, there can be more than one
\newglossary{symbols}{lds}{lsi}{Symbols List}
\newglossary{index}{ing}{ige}{General Index}

%setting external documents and prefixes
\newcommand{\setfileprefix}[2]{\csdef{prefix@#1}{#2}}
\newcommand{\usefileprefix}[1]{\csuse{prefix@#1}}

\setfileprefix{MyDoc1.pdf}{II-}
\setfileprefix{MyDoc2.pdf}{III-}

\renewcommand*\glsxtrdisplaysupploc[5]{%
  \setentrycounter[#1]{#2}%
  \usefileprefix{#4}\glsxtrmultisupplocation{#5}{#4}{#3}%
}

%style of symbols list
\newglossarystyle{listsimb}{%
\glossarystyle{super}%
\renewenvironment{theglossary}%
    {\tablehead{}\tabletail{}%
     \begin{supertabular}{@{}lp{20cm}}}%
    {\end{supertabular}}%
\renewcommand{\glsgroupskip}{}%
\renewcommand*{\glossaryentryfield}[5]{%
\glsentryitem{##1}\glstarget{##1}{##2} & ##3\glspostdescription\space ##5\\[2pt]}%
}

%font
\renewcommand*{\glsnamefont}[1]{\textmd{#1}}

%loading supplemental files
\GlsXtrLoadResources[
 src=index,
 type=index,
 supplemental-locations={MyDoc1, MyDoc2},% you can have more than one!
 supplemental-category={supplemental},
 supplemental-selection={all}
]

%and you can define all kinds of glossaries
\GlsXtrLoadResources[
 src=symbols,
 type=symbols,
 supplemental-locations={MyDoc1, MyDoc2},
 supplemental-category={supplemental},
 supplemental-selection={all}
]

\begin{document}
A test \gls{ind2} and \gls{symb2}.

\printunsrtglossary[type=index, style=bookindex]
\printunsrtglossary[type=symbols, style=listsimb]
\end{document}

You will also need files for index and symbols in this example index.bib

@index{ind1, %id
 name = {Something}, %as is going to appear in the index
 user1 = {some-thing} %if you want something different in the text
}

  @index{ind2,
     name = {another thing},
     parent = {ind1}
  }

And symbols.bib

@entry{symb1,
 name = {\ensuremath{c}},
 description = {an important constant},
 sort = {constant}
}

@entry{symb2,
 name = {\ensuremath{\binom{n}{k}}},
 description = {binomial coeficient},
 sort = {binomial}
}

All of them can have a sort key, but in the index case it is correct.

Then, compile with the following chain:

pdflatex MyDoc1
pdflatex MyDoc2
pdflatex MyDoc3
bib2gls MyDoc1
bib2gls MyDoc2
bib2gls MyDoc3 
pdflatex MyDoc1
pdflatex MyDoc2
pdflatex MyDoc3

and you will get the following in the different files