esindex (30sep19)

git-svn-id: svn://tug.org/texlive/trunk@52226 c570f23f-e606-0410-a88d-b1316a301751

4 files changed, 155 insertions, 43 deletions

diff --git a/Master/texmf-dist/doc/latex/esindex/README.md b/Master/texmf-dist/doc/latex/esindex/README.md
index df626eab5fc..a2b357cd576 100644
--- a/Master/texmf-dist/doc/latex/esindex/README.md
+++ b/Master/texmf-dist/doc/latex/esindex/README.md
@@ -1,8 +1,35 @@
+## esindex
 
 Generates automatically sort keys. Originally only for Spanish, but
 version 1.5 provides tools for it to be adapted to other languages, so
-that, for example, \esindex{\textit{Höhe}} can be made equivalent to
+that, for example,
+```
+\esindex{\textit{Höhe}}
+```
+can be made equivalent to
+```
 \index{Hoehe@\textit{Höhe}}
+```
+
+This package is based on the principle that sort keys should be
+generated in most cases in the context where the entry appears
+in the document.
+
+### Recent changes
+
+```
+1.6   2019-09-30
+      - Package option 'babel', which wraps each entry not in the main
+        language with \foreignlanguage (or other macro).
+
+1.5   2019-01-28
+      - Macro \esindexactual, to replace all ocurrences of an entry
+        with a new text.
+      - Package option 'nospanish', for it to be adapted to other
+        languages.
+      - \esindexreplace, \esindexlastchar, \esindexkey now work with
+        xe/lua.
+``` 
 
 License:     MIT
 ________
diff --git a/Master/texmf-dist/doc/latex/esindex/esindex.pdf b/Master/texmf-dist/doc/latex/esindex/esindex.pdf
index f94fd894349..d1e0de5e5f2 100644
--- a/Master/texmf-dist/doc/latex/esindex/esindex.pdf
+++ b/Master/texmf-dist/doc/latex/esindex/esindex.pdf
diff --git a/Master/texmf-dist/doc/latex/esindex/esindex.tex b/Master/texmf-dist/doc/latex/esindex/esindex.tex
index 467d14ed4df..a654b2aa8d5 100644
--- a/Master/texmf-dist/doc/latex/esindex/esindex.tex
+++ b/Master/texmf-dist/doc/latex/esindex/esindex.tex
@@ -7,31 +7,60 @@
 % Repository: https://github.com/jbezos/esindex
 %
 
-\documentclass{article}
+\documentclass[a4paper]{ltxguide}
+
 \usepackage[english,spanish]{babel}
 \spanishdatedel
 \usepackage[utf8]{inputenc}
 \usepackage[T1]{fontenc}
 
-\title{Paquete \textsf{esindex}\footnote{Este
-     archivo est\'a actualmente en la versión 1.5
-     con fecha 2019-01-28.
-     Esta copia del manual se compuso el~\today.}}
+\title{\textsf{esindex}\\\large 1.6\quad 2019-09-30}
 
-\author{Javier Bezos\footnote{Para comentarios y sugerencias: 
-\texttt{http://www.texnia.com}.}}
+\author{Javier Bezos}
 
 \raggedright
-\parskip=1ex
-
-\date{2019-01-28}
+\parskip=.8ex
+\advance\oddsidemargin-.7cm
+\advance\textwidth1.8cm
+\addtolength{\textheight}{3.5cm}
+\addtolength{\topmargin}{-2cm}
+
+\usepackage{xcolor,bera}
+
+\definecolor{notes}{rgb}{.75, .3, .3}%
+
+\makeatletter
+\def\@begintheorem#1#2{%
+  \list{}{}%
+  \global\advance\@listdepth\m@ne
+  \item[{\sffamily\bfseries\color{notes}\MakeUppercase{#1}}]}%
+\makeatother
+\newtheorem{warning}{Warning}
+\newtheorem{note}{Note}
+\newtheorem{example}{Example}
 
 \begin{document}
 
-\maketitle
+\vspace*{1cm}
+{\fontsize{48}{48}\selectfont \color{notes}{esindex}\par}
+{\LARGE Generating automatically sort keys for \textit{MakeIndex}
+with \LaTeX\par}
+\vspace*{1ex}
+Version 1.6 (2019-09-30)\par
+Javier Bezos (\texttt{http://www.texnia.com})
+
+\vspace*{6ex}
+
+{\small\itshape Please, report any issues you find in
+  \texttt{https://github.com/jbezos/esindex/issues}.\par}
+
+This package defines the command \verb|\esindex|, which adds sort keys
+to the index entries for you. Although originally a specifically Spanish
+tool (so part of the documentation is in Spanish), the idea behind this
+package may be applied to other languages, and tools to adapt it are
+provided. They are explained below, section 2.
 
-This package defines the command \verb|\esindex| which eases writing
-Spanish index entries:
+So,
 \begin{verbatim}
 \esindex{cañón}
 \end{verbatim}
@@ -39,17 +68,24 @@ is equivalent to
 \begin{verbatim}
 \index{can^^ffon@cañón}
 \end{verbatim}
-As you can see, the package generates the sort key within \TeX{}
-itself, which has a number of advantages.
 
-Although it is a specifically Spanish tool (so most of the
-documentation is in Spanish), the idea behind this package may be
-applied to other languages, and tools to adapt it are provided, too.
-They are explained below.
+\verb|\index| is not touched at all (except if requested with the package
+option \verb|replaceindex|).
+
+As you can see, the package generates the sort key within \TeX{}
+itself, which has a number of advantages (and some disadvantages, of
+course).
 
 Version 1.5 provides tools for other languages and adds support for
 \textsf{luatex} and \textsf{xetex}.
 
+Version 1.6 adds a package option -- with \verb|babel| the `actual' value
+(ie, the text to be printed in the index) is wrapped with
+\verb|\foreignlanguage| if the current language is not the main one
+(this replacement comes after \verb|\esindexactual|). Actually, the
+macro is \verb|\esindexlanguage|, which by default is `let' to the
+\textsf{babel} macro.
+
 \section{Spanish}
 
 Este paquete ha sido diseñado para facilitar la escritura de índices
@@ -151,9 +187,9 @@ entradas idénticas.
 
 \item La orden \verb|\esindexsort| permite predefinir claves asociadas
 a entradas concretas, para ajustar su ordenación (lo que normalmente se
-consigue añadiendo texto adicional para que makeindex lo tenga en
-cuenta). Estas correspondencias deben darse antes de la aparición del
-primer \verb|\esindex| con ese término, y las claves se procesan
+consigue añadiendo texto adicional para que \textsf{makeindex} lo tenga
+en cuenta). Estas correspondencias deben darse antes de la aparición
+del primer \verb|\esindex| con ese término, y las claves se procesan
 posteriormente con \verb|ignorespaces|, \verb|\ignorewords| y
 \verb|\everyesindex|, si están activadas. Por ejemplo:
 \begin{verbatim}
@@ -206,7 +242,7 @@ is still based on “Felipe II”, of course).
 
 \end{enumerate}
 
-Do you find it chaotic? Well, you are right. After all this package is
+Do you find it chaotic? Well, you are right. After all this package was
 for Spanish indexes with some readjustments for it to be adapted to
 other languages. A general solution deserves another package. Feel free
 to create one based on this package (MIT license), if you like.
@@ -237,14 +273,62 @@ for this precise expansion.
 \item Make a direct replacement of some characters.
 
 \end{enumerate}
+Note there is a pattern: first flatten the key by removing macros, and
+then the resulting unmarked text is processed to replace some chars.
+
 Remember these changes are not shown -- they are used by
 \textit{MakeIndex} to sort the entries. Note also replacements can be
-language dependent with the appropriate test (eg, with
+made language dependent with the appropriate test (eg, with
 \textsf{iflang}).
 
 As a convenience tool, \verb|\esindexlastchar| is \verb|^^ff| or
 \verb|^^^^ffff|, depending on the engine.
 
+You can play with \textsf{lua} inside \verb|\everyesindex|, too. For
+example, to convert at once things like \textit{2,3-dimetilfenol} to
+\textit{dimetilfenol}:
+\begin{verbatim}
+\everyesindex{
+  \directlua{
+    token.set_macro('esindexkey',
+                    string.gsub([[\esindexkey]], '^[-0-9,]+', ''))
+  }}
+\end{verbatim}
+
+It is just the concept (for example, it fails if the key contains
+\verb|]]|).
+
+\section{Final remarks}
+
+Indexing has many subtleties. Recommended readings are:
+\begin{itemize}
+  \item \textit{Guidelines for Alphabetical Arrangement of Letters and
+  Sorting of Numerals and Other Symbols},
+  \verb|https://www.niso.org/sites/default/files/2017-08/tr03.pdf|
+  \item \textit{Guidelines for Indexes and Related Information Retrieval
+  Devices},
+  \verb|https://www.niso.org/sites/default/files/2017-08/tr02.pdf|
+  \item \textit{Unicode Collation Algorithm},
+  \verb|https://unicode.org/reports/tr10/|,
+  \item \textit{Unicode Locale Data Markup Language (LDML)}, part 5,
+  \textit{Collation},
+  \verb|http://www.unicode.org/reports/tr35/tr35-collation.html|.
+\end{itemize}
+
+There are at least 3 tasks, from \LaTeX{} to a suitable key (eg,
+\verb|$\delta$| to \verb|delta| or whatever you want), collating these
+keys, and generate the formated index in a form \LaTeX{}
+understands.\footnote{There can be also sub-tasks. For example, the
+last task is best split in two sub-tasks, namely, first generate a
+sorted list of terms, and then format them, but unfortunately, most
+\texttt{ist} styles output a formatted list. We also need to format and
+collapse locators and so on.} With \textsf{makeindex} these 3 steps are
+intermingled, and so does \textsf{esindex}. This is another reason why
+a general solution deserves a completely new package. However, with
+some patience you can do many useful things with \textsf{esindex}, even
+sorting words in non-Latin scripts (at least, with the most basic
+rules).
+
 \end{document}
 
 MIT License
diff --git a/Master/texmf-dist/tex/latex/esindex/esindex.sty b/Master/texmf-dist/tex/latex/esindex/esindex.sty
index ac00170d3a8..4594874db47 100644
--- a/Master/texmf-dist/tex/latex/esindex/esindex.sty
+++ b/Master/texmf-dist/tex/latex/esindex/esindex.sty
@@ -6,21 +6,8 @@
 %
 % Repository: https://github.com/jbezos/esindex
 %
-% Changes
-%
-% 2006-05-12. Bug fixed. Missing accents in multilevel entries.
-% 2008-09-17. Made compatible with latest LaTeX versions (\IeC).
-%             Added \'\i.
-%             Package option ignorespaces.
-%             Macro \ignorewords.
-% 2008-09-28. Added \everyesindex.
-%             Macro \esindexsort.
-% 2019-01-25. Macro \esindexactual.
-%             Package option nospanish.
-%             Adapted to xe/lua, with:
-%             Macros \esindexreplace, \esindexlastchar, \esindexkey
-
-\ProvidesPackage{esindex}[2019/01/28 v1.5 Spanish indexes]
+
+\ProvidesPackage{esindex}[2019/09/30 v1.6 Spanish indexes (and more)]
 
 \def\esx@level{`\!}
 \def\esx@actual{`\@}
@@ -36,6 +23,12 @@
 
 \DeclareOption{nospanish}{\esx@spanishfalse}
 
+\newif\ifesx@babel
+
+\DeclareOption{babel}{%
+  \let\esindexlanguage\foreignlanguage
+  \esx@babeltrue}
+
 \DeclareOption*{%
   \@expandtwoargs\in@{=}{\CurrentOption}%
   \ifin@
@@ -50,8 +43,8 @@
     {\string`\expandafter\noexpand\csname\string#2\endcsname}}}
 
 \def\esx@error{\PackageError{esindex}{Unknown index parameter}%
-	   {Except for ignorespaces, nospanish, replaceindex, options^^J%
-     in this package ave the form parameter=value, where^^J%
+	   {Except for ignorespaces, nospanish, babel, replaceindex, ^^J%
+     options in this package ave the form parameter=value, where^^J%
      parameter stands for of the following MakeIndex ones:^^J%
      encap, level, actual, quote}}
 
@@ -141,9 +134,17 @@
       % Set actual, if there is a saved one with \esindexactual.
   \def\esx@a{#1}%
   \@ifundefined{esx@actual@\expandafter\strip@prefix\meaning\esx@a}%
-    {\def\esx@a{#1}}%
+    {}%
     {\expandafter\let\expandafter\esx@a
        \csname esx@actual@\expandafter\strip@prefix\meaning\esx@a\endcsname}%
+      % Wrap with the language, if not main (babel option). This idea
+      % can be extended to consider more generic contexts
+  \ifesx@babel
+    \bbl@ifsamestring\languagename\bbl@main@language
+      {}%
+      {\@temptokena\expandafter{\esx@a}%
+       \edef\esx@a{\string\esindexlanguage{\languagename}{\the\@temptokena}}}%
+  \fi
       % Add actual, and go to the next ! level
   \toks@\expandafter{\esx@g}%
   \def\esx@b{#2}%