French spacing without babel or polyglossia

Question

To re-edit a document in a constructed language (latino sine flexione), I need "French" punctuation spacing without side effects.

More specifically, I want a small non-breaking space before any double punctuation marks -- : ; ! ? -- and a regular non-breaking space after an opening or before a closing quotation mark (guillemet) -- « and ».

Answer 1

For pdflatex you can use solution based on changing category code of characters:

\documentclass{article}
\usepackage[utf8]{inputenc}
\usepackage[T1]{fontenc}
% saving original quotes representation
\let\gll\guillemotleft
\let\glr\guillemotright
% new quotes representation
\def\frenchguillemotleft{\gll~}
\def\frenchguillemotright{~\glr}
{
    \catcode`\:=\active
    \catcode`\;=\active
    \catcode`\?=\active
    \catcode`\!=\active
    \gdef\frenchpunct{%
        \catcode`\:=\active \def:{\thinspace\char`\: }
        \catcode`\;=\active \def;{\thinspace\char`\; }
        \catcode`\?=\active \def?{\thinspace\char`\? }
        \catcode`\!=\active \def!{\thinspace\char`\! }
        \let\guillemotleft\frenchguillemotleft
        \let\guillemotright\frenchguillemotright
    }
}
\def\nofrenchpunct{%
    \catcode`\:=12
    \catcode`\;=12
    \catcode`\?=12
    \catcode`\!=12
    \let\guillemotleft\gll
    \let\guillemotright\glr
}
\begin{document}
french: french; french? french!
«french»

\frenchpunct
french: french; french? french!
«french»

\nofrenchpunct
french: french; french? french!
«french»
\end{document}

Command \frenchpunct make characters :, ;, ? and ! active and then define them according to desired behaviour: unbreakable thin space before character. Also this command redefine commands \guillemotleft and \guillemotright those are declared by inputenc package.

Command \nofrencpunct change all back.

Remark: the above \frenchpunct command is local. Therefore being enclosed by groping characters {} it take off its effect after leaving group. If you need global command that is canceled only by \nofrenchpunct add \global before all \def (or use \gdef that is alias for \global\def), \catcode and \let in \frenchpunct definition:

\gdef\frenchpunct{%
    \global\catcode`\:=\active \gdef:{\thinspace\char`\: }
    \global\catcode`\;=\active \gdef;{\thinspace\char`\; }
    \global\catcode`\?=\active \gdef?{\thinspace\char`\? }
    \global\catcode`\!=\active \gdef!{\thinspace\char`\! }
    \global\let\guillemotleft\frenchguillemotleft
    \global\let\guillemotright\frenchguillemotright
}

And make same for \nofrenchpunct:

\def\nofrenchpunct{%
    \global\catcode`\:=12
    \global\catcode`\;=12
    \global\catcode`\?=12
    \global\catcode`\!=12
    \global\let\guillemotleft\gll
    \global\let\guillemotright\glr
}

Answer 2

Here's a LuaLaTeX-based solution. It consists of a Lua function called french_punctuation_spacing that does all of the work and two LaTeX utility macros called \FrenchPunctuationSpacingOn and \FrenchPunctuationSpacingOff, respectively, which activate and deactivate the Lua function.

% !TEX TS-program = lualatex
\documentclass{article}
\usepackage{luacode}  % for 'luacode' environment
\begin{luacode}

function french_punctuation_spacing ( s )
  s = s:gsub ( "[:;!?]" , "\\,%0" )
  s = s:gsub ( "«" , "«~" )
  s = s:gsub ( "»" , "~»" )
  return s
end

\end{luacode}
\newcommand\FrenchPunctuationSpacingOn{\directlua{%
  luatexbase.add_to_callback ( "process_input_buffer",
  french_punctuation_spacing , "frenchpunctuationspacing" )}}
\newcommand\FrenchPunctuationSpacingOff{\directlua{%
  luatexbase.remove_from_callback ( "process_input_buffer",
  "frenchpunctuationspacing" )}}

%\usepackage{fontspec} % optional

\begin{document}
bonjour: monde; oui? non! «aujourd'hui»

\FrenchPunctuationSpacingOn % activate the Lua function
bonjour: monde; oui? non! «aujourd'hui»

\FrenchPunctuationSpacingOff % deactivate the Lua function
bonjour: monde; oui? non! «aujourd'hui»
\end{document}