\documentclass[12pt]{ltxdoc}
%\documentclass{article}
%\usepackage[utf8]{luainputenc}
\usepackage{fontspec}
\setmainfont[Ligatures=TeX]{Linux Libertine O}
\usepackage[czech,english]{babel}
\usepackage{luacode}
\usepackage{amsfonts}
%\usepackage[]{polyglossia}
%\setmainlanguage{english}
%\setotherlanguage{czech}
%\usepackage{url}
\usepackage{hyperref}
%\input prevent-single
%\usepackage[]{prevent-single}
\usepackage[]{luavlna}
%\input prevent-single
%\def\preventsingledebugon{}
\newenvironment{mycode}{
	%\leavevmode%
	\medskip%
	\parindent=0pt%
}{\medskip}

\ifdefined\version\else
  \def\version{{{VERSION}}}
  \let\gitdate\date
\fi

\begin{document}
\title{The \verb|luavlna| package}
\author{Michal Hoftich (\url{michal.h21@gmail.com})\\Miro Hrončok (\url{miro@hroncok.cz})}
\date{Version \version\\\gitdate}
\maketitle
\tableofcontents
\section{Introduction}


This is a small package for Plain Lua\TeX and Lua\LaTeX. In some languages, like
Czech or Polish, there should be no single letter words at the line
end, according to the typographical norms. There exists some
external commands (like \verb!vlna!) or packages (\verb!encxvlna!
for enc\TeX, \verb!xevlna! for Xe\TeX,
\verb!impnattypo! for Lua\LaTeX). %This package is for plain
%lua\TeX and for lua\LaTeX.

Another feature is inclusion of a non-breakable space after 
initials, like in personal names, after or before academic degrees
and between numbers and units (SI and others).

The code is modified version of Patrick Gundlach's answer on
TeX.sx\footnote{\url{http://tex.stackexchange.com/a/28128/2891}}.
The difference is that it is possible to specify which single letters
should be taken into account for different languages.
% The support for degrees and units was added as well.
% The code works also for single letters at the beginning of the brackets.



\section{Usage}

The usage is simple:

\begin{verbatim}
\input ucode
\uselanguage{czech}
%% in the case of luacsplain, use instead:
%% \chyph
\input luavlna
\preventsingledebugon
\input luaotfload.sty
\font\hello={name:Linux Libertine O:+rlig;+clig;+liga;+tlig} at 12pt 
\hsize=3in
\hello
Příliš žluťoučký kůň úpěl ďábelské ódy. 
Text s krátkými souhláskami a samohláskami i dalšími jevy 
z nabídky možností (v textu možnými). 

I začátek odstavce je třeba řešit, i když výskyt zalomení 
není pravděpodobný.

Co třeba í znaky š diakritikou?

Různé možnosti [v závorkách <i jiných znacích

Podpora iniciál a titulů: M. J. Hegel, Ing. Běháková, Ph.D., Ž. Zíbrt, 
Ch. Borner.

Podpora jednotek: 100,5 MN\cdot{}s, 100.5 kJ, 200 µA, $-1$ dag, 
12 MiB.

Uvnitř matematiky by mělo být zpracování vypnuté:  $k \in \mathbb N$. 
Pokračujeme v příkladu.

\preventsingledebugoff
\bye
\end{verbatim}

%\noindent\parbox{3in}{%

\noindent
\begin{minipage}{3in}
\preventsingledebugon
\selectlanguage{czech}
Příliš \textit{žluťoučký kůň} úpěl ďábelské ódy. 
Text s krátkými souhláskami a samohláskami i dalšími jevy z nabídky možností (v textu možnými). 

I začátek odstavce je třeba řešit, i když výskyt zalomení není pravděpodobný.

Co třeba í znaky š diakritikou?

Různé možnosti [v závorkách \textless i jiných znacích

Podpora iniciál a titulů: M. J. Hegel, Ing. Běháková, Ph.D., Ž. Zíbrt, Ch. Borner.

Podpora jednotek: 100,5 MN\cdot{}s, 100.5 kJ, 200 µA, $-1$ dag, 12 MiB.

Uvnitř matematiky by mělo být zpracování vypnuté:  $k \in \mathbb N$. Pokračujeme v příkladu.

\preventsingledebugoff
%}
\end{minipage}

\selectlanguage{english}
\bigskip
It is also possible to use the package with Lua\LaTeX, just use

\begin{verbatim}
	\usepackage{luavlna}
\end{verbatim}

in the preamble.

\section{Package options}
The \verb|luavlna| package recognizes the following options

\begin{description}
  \item[\texttt{noprocess}] -- disable the automatic document processing.
  \item[\texttt{noinitials}] -- disable processing of the initials.
  \item[\texttt{nounits}] -- disable processing of the SI units.
  \item[\texttt{nopredegrees}] -- disable processing of the the academic degrees before names.
  \item[\texttt{nosufdegrees}] -- disable processing of the the academic degrees after names.
  \item[\texttt{debug}] -- enable the debugging mode.
  \item[\texttt{defaults}] -- load default settings. Enabled by default.
\end{description}


\section{Default values}

Luavlna defines default settings for Czech and Slovak. If you want to disable them,
you can use the \texttt{defaults=false} option for Luavlna package. In Plain Lua\TeX{},
use:

\begin{mycode}
\begin{verbatim}
\let\nosingledefaults\relax
\input luavlna
\end{verbatim}
\end{mycode}

\section{Commands}

\begin{mycode}
\cmd{\singlechars}\marg{language name}\marg{letters} 
\end{mycode}

Enable this feature for certain letters in given language. 
%Language code is internall \TeX\ code for the language, it is $0$ for English, 
%$16$ for Czech. Please note that in \verb|csplain|, language code for Czech 
%is $5$ and you will have to set it yourself.

Default values:

\begin{mycode}
\begin{verbatim}
%% only Czech and Slovak are supported out of the box
\singlechars{czech}{AIiVvOoUuSsZzKk}
\singlechars{slovak}{AIiVvOoUuSsZzKk}
\end{verbatim}
\end{mycode}


%By default, all
%single letters are processed, this command can be used to pass a
%string of characters, which should be processed only.

\begin{mycode}
\cmd{\compoundinitials}\marg{language name}\marg{compounds}
\end{mycode}

Declare compound letters for given language. Second argument should be comma 
separated list of compound letters, in exact form in which they can appear.

Default values:

\begin{mycode}
\begin{verbatim}
\compoundinitials{czech}{Ch,CH}
\end{verbatim}
\end{mycode}


\subsection{Split hyphens}

The hyphenated words like ``je-li'' should be be hyphenated as ``je-/-li``
according to the Czech typesetting rules. This behaviour can be enabled or disabled for a particular
language using the following commands:

\begin{mycode}
  \cmd{\enablesplithyphens}\marg{language name}\\
  \cmd{\disablesplithyphens}\marg{language name}
\end{mycode}

By default, it is enabled for the Czech language.

Example in action:

\begin{minipage}{3in}
  \selectlanguage{czech}
   Sedlec-Prčice, modro-zelený,  překladatel-tlumočník, kuchař-číšník, propan-butan,                                                                                          
   Otýlie Sklenářová-Malá, František Jílek-Oberpfalcer.
  \selectlanguage{english}
\end{minipage}

The similar behaviour can be achieved using the \verb|\splithyphens|  command
provided by Czech langauge definitions for the Babel package, but it's use is
discouraged. It prevents use ot the hyphen character in many situations. The
solution provided by Luavlna should be safe.


\subsection{Turning off language switching}

By default, language of the nodes is taken into account. If you want to use
settings for one language for a whole document, you can use following command:

\begin{mycode}
\cmd{\preventsinglelang}\marg{language name}
\end{mycode}

\subsection{Turning off processing}

If you want to stop the processing of spaces in the document you can use the following command:

\begin{mycode}
\cmd{\preventsingleoff}
\end{mycode}

To resume processing, use

\begin{mycode}
\cmd{\preventsingleon}
\end{mycode}

You can also disable initials, units and degrees processing:

\begin{mycode}
  \cmd{\noinitials}\\
  \cmd{\nounits}\\
  \cmd{\nopredegrees}\\
  \cmd{\nosufdegrees}
\end{mycode}


\subsection{Debugging commands}
\begin{mycode}
\cmd{\preventsingledebugon}\par
\cmd{\preventsingledebugoff}
\end{mycode}

Insert debugging marks on/off. Default off.

\section{Lua module \texttt{luavlna-langno.lua}}

When we process glyph nodes with Lua\TeX\ callbacks, there are \verb|lang| 
fields available. These are numerical codes of languages, but no information
about language names easily accesible from Lua side is available.\footnote{%
	Language names are stored in \TeX macros like \verb|\string\l@langname|, but 
	different formats use different naming of these macros}
These numbers are format dependent, majority of formats like 
Lua\LaTeX use \verb|language.dat| file provided by \verb|babel| 
for assign numbers to languages, but for example \verb|csplain| 
use its own system.

To allow easy setting of language dependent parameters, \verb|langno| module 
was created. It's purpose is to translate language code to language name and 
the other way. Lua\LaTeX, Lua\TeX\ and CSplain formats are supported at the moment.

\subsection{Recognized languages}
\subsubsection{Lua\TeX\ and Lua\LaTeX}

File \verb|language.dat| is processed to load language names, aliases and assigned numbers. These language names are the same as names supported by the \verb|babel| package.

\begin{quotation}
  \small\noindent
\begin{luacode*}
  langno = require "luavlna-langno"
  function print_format_lang(fmt)
    local luatex = langno.load_languages(fmt)
    local t = {}
    for k, v in pairs(luatex.names) do
      if k and k~="" then
        t[#t+1]=k
      end
    end
    table.sort(t)
    tex.print(table.concat(t,", "))
  end
  print_format_lang("luatex")
\end{luacode*}
\end{quotation}

\subsubsection{CSplain}

Different method is used with the CSplain format. The \verb|lua-hyphen.lan| file is parsed for the language codes.

\begin{quotation}
  \small\noindent
  \begin{luacode*}
    print_format_lang("csplain")
  \end{luacode*}
\end{quotation}


\section{License}

Permission is granted to copy, distribute and/or modify this software
under the terms of the LaTeX Project Public License, version 1.3.

\end{document}