path: root/Master/texmf-dist/source/latex/units/units.dtx

%    \iffalse
%
% units.dtx
% Docstrip archive, run through LaTeX.
%
% Copyright (C) 1998 Axel Reichert
% See the files README and COPYING.
%
%    \fi
%
%    \CheckSum{174}
%
%    \changes{v0.9a}{1998-06-25}{First release}
%    \changes{v0.9b}{1998-08-04}{Documentation update}
%
%% \CharacterTable
%%  {Upper-case    \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
%%   Lower-case    \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
%%   Digits        \0\1\2\3\4\5\6\7\8\9
%%   Exclamation   \!     Double quote  \"     Hash (number) \#
%%   Dollar        \$     Percent       \%     Ampersand     \&
%%   Acute accent  \'     Left paren    \(     Right paren   \)
%%   Asterisk      \*     Plus          \+     Comma         \,
%%   Minus         \-     Point         \.     Solidus       \/
%%   Colon         \:     Semicolon     \;     Less than     \<
%%   Equals        \=     Greater than  \>     Question mark \?
%%   Commercial at \@     Left bracket  \[     Backslash     \\
%%   Right bracket \]     Circumflex    \^     Underscore    \_
%%   Grave accent  \`     Left brace    \{     Vertical bar  \|
%%   Right brace   \}     Tilde         \~}
%
%    \newcommand{\file}[1]{\texttt{#1}}
%    \newcommand{\option}[1]{\texttt{#1}}
%    \newcommand{\package}[1]{\texttt{#1}}
%    \newcommand{\person}[1]{\textsc{#1}}
%    \newcommand{\winkey}{\texttt{(-;}}
%
%    \title{\file{units.sty} -- \file{nicefrac.sty}}
%    \author{Axel Reichert}
%    \date{1998-08-04}
%    \maketitle
%    \begin{abstract}
%      \file{units.sty} is a package for setting units in a
%      typographically correct way.  It is based upon
%      \file{nicefrac.sty}, a package for nice fractions.  See the
%      files \file{README} and \file{COPYING} for additional
%      information.
%    \end{abstract}
%    \section{Loading}
%    \begin{tabbing}
%      Only nice fractions: \= |\usepackage{nicefrac}| \\
%      Only units or both:  \> |\usepackage{units}|
%    \end{tabbing}
%    \section{Options}
%    \begin{tabbing}
%      Tight spacing for units (default): \=
%        |\usepackage[tight]{units}| \\
%      Loose spacing for units: \>
%        |\usepackage[loose]{units}| \\
%      ``Nice'' fractions (default): \>
%        |\usepackage[nice]{nicefrac}| \\
%      ``Ugly'' fractions: \>
%        |\usepackage[ugly]{nicefrac}|
%    \end{tabbing}
%
%    The options \option{nice} and \option{ugly} can also be used for
%    the \package{units} package, they will simply be passed to the
%    \package{nicefrac} package, so you can combine the options,
%    e.\,g.: |\usepackage[loose,ugly]{units}|
%
%    Tight spacing means \cmd{\,} for the space between the value and the
%    dimension, loose spacing uses |~|, like \unit[1]{m} and 1~m.  Nice
%    fractions look like \nicefrac{m}{s}, ugly fractions look like m/s
%    in text mode and $\frac{\mathrm{m}}{\mathrm{s}}$ in math mode.
%    \section{Commands}
%    \begin{tabbing}
%      Fractions of units: \=
%        |\unitfrac[|\meta{val}|]{|\meta{num}|}{|\meta{denom}|}| \kill
%      Units: \> |\unit[|\meta{val}|]{|\meta{dim}|}| \\
%      Fractions of units: \>
%        |\unitfrac[|\meta{val}|]{|\meta{num}|}{|\meta{denom}|}| \\
%      Nice fractions: \>
%        |\nicefrac[|\meta{fontcmd}|]{|\meta{num}|}{|\meta{denom}|}|
%    \end{tabbing}
%    In these list, \meta{val} and \meta{dim} denote the value and the
%    dimension of the unit, respectively.  \meta{num} and \meta{denom}
%    are the numerator and denominator of the fraction, and
%    \meta{fontcmd} can be an author command for fonts or a math
%    alphabet, see \file{fntguide.dvi}.
%
%    Typically, \meta{val} is only a number and \meta{num} and
%    \meta{denom} are relatively simple \LaTeX{} expressions.  If you
%    really feel the need for putting whole paragraphs or complex
%    formula into such a fraction, you have misunderstood the purpose
%    of this package.  \winkey
%
%    A common mistake is to forget the enclosing braces when using the
%    optional argument of the commands inside another optional
%    argument.  This is no package bug, but a \LaTeX{} feature.  Correct
%    example:  |\section[{\unit[1]{m}}]{...}|.\footnote{Credits go to
%    \person{Frank Knappe} for suggesting this documentation update.}  
%
%    It is very important to be aware of the fact that all these
%    commands distinguish between text mode and math mode.  Within text
%    mode, the font of the surrounding text will be used by default,
%    for math mode \cmd{\mathrm} is the default.
%
%    This is quite sensible, because you would not want your
%    collection of delicious recipes (typeset in a mega-cool ultra
%    condensed bold italic calligraphical font) contain those spindle
%    Computer Modern Roman just for half a litre of milk.  \winkey
%
%    Otherwise, when working on scientific papers, strict and
%    consistent notation is really a virtue, and, like it or not,
%    units are typeset with upright fonts.  So take great care when
%    deciding about math mode or not.
%    \section{Examples}
%    \begin{tabbing}
%      |\sffamily\itshape$\unitfrac[1]{m}{s}$| \quad \=
%        \sffamily\itshape$\unitfrac[1]{m}{s}$ \kill
%      |\sffamily\bfseries\unit{m}| \> \sffamily\bfseries\unit{m} \\
%      |\sffamily\bfseries$\unit{m}$| \> \sffamily\bfseries$\unit{m}$ \\
%      |\sffamily\itshape\unit[1]{m}| \> \sffamily\itshape\unit[1]{m} \\
%      |\sffamily\itshape$\unit[1]{m}$| \> \sffamily\itshape$\unit[1]{m}$ \\
%      |\sffamily\bfseries\unitfrac{m}{s}| \>
%        \sffamily\bfseries\unitfrac{m}{s} \\
%      |\sffamily\bfseries$\unitfrac{m}{s}$| \>
%        \sffamily\bfseries$\unitfrac{m}{s}$ \\
%      |\sffamily\itshape\unitfrac[1]{m}{s}| \>
%        \sffamily\itshape\unitfrac[1]{m}{s} \\
%      |\sffamily\itshape$\unitfrac[1]{m}{s}$| \>
%        \sffamily\itshape$\unitfrac[1]{m}{s}$
%    \end{tabbing}
%    As you can see, font changes are ignored in math mode~\ldots
%    \begin{tabbing}
%      |\scriptsize\sffamily\itshape$\unitfrac[1]{m}{s}$| \quad \=
%        \scriptsize\sffamily\itshape$\unitfrac[1]{m}{s}$ \kill
%      |\scriptsize\sffamily\itshape\unitfrac[1]{m}{s}| \>
%        \scriptsize\sffamily\itshape\unitfrac[1]{m}{s} \\
%      |\scriptsize\sffamily\itshape$\unitfrac[1]{m}{s}$| \>
%        \scriptsize\sffamily\itshape$\unitfrac[1]{m}{s}$
%    \end{tabbing}
%    \ldots~except for the fontsize.
%    \begin{tabbing}
%      |\bfseries\itshape$\nicefrac{1}{2}$| \quad \=
%        \bfseries\itshape$\nicefrac{1}{2}$ \kill
%      |\bfseries\itshape\nicefrac{1}{2}| \>
%        \bfseries\itshape\nicefrac{1}{2} \\
%      |\bfseries\itshape$\nicefrac{1}{2}$| \>
%        \bfseries\itshape$\nicefrac{1}{2}$ \\
%      |\nicefrac[\texttt]{1}{2}| \> \nicefrac[\texttt]{1}{2} \\
%      |\nicefrac[\texttt]{\textit{1}}{2}| \>
%        \nicefrac[\texttt]{\textit{1}}{2} \\
%      |$\nicefrac[\mathcal]{A}{B}$| \> $\nicefrac[\mathcal]{A}{B}$
%    \end{tabbing}
%    The \cmd{\nicefrac} command can deal even with quite strange font
%    changing commands.
%    \section{Typography}
%    Why should units be typeset in upright fonts, not in italics?
%    Because they have to be distinguished from normal variables:
%    ``\unit{m}'' is meter, ``$m$'' is a variable, for example a mass.
%
%    Why should the space between the value and the dimension be
%    non-breakable?  Because the reader is disturbed by linebreaks like
%    1 $\mathrm{m}$.
%
%    Why should the space between the value and the dimension be a
%    half word space only?  Because things belonging together are
%    typeset tighter.  Compare \unit[1]{m} with 1~m and the normal word
%    spacing, which can vary from line to line.
%
%    Why should nice fractions be typeset so that the numerator does
%    not extend above the height of the letter ``M'' and the
%    denominator does not extend below the baseline?  Because otherwise
%    a stretching of the baselineskip could be necessary due to
%    descenders.
%
%    By the way, a very common mistake is to place units into
%    brackets, like [\unit{N}].  The correct notation is $[F] =
%    \unit{N}$.  The brackets indeed are a function with a variable as
%    an argument.  The value returned is the unit.  If you need to
%    specify units in table headings, then use a single row for the
%    units, where you put them into \emph{parentheses} instead.
%    \section{Bugs}
%    None, are you kidding?
%    \section{Features}
%    \begin{itemize}
%    \item Consistent and logical markup of units is enhanced instead
%      of fiddling around with spacing and fonts.
%    \item The same command works in text mode \emph{and} math mode,
%      font and size are adjusted automatically for nice integration
%      within text mode while forcing strict notation in math mode.
%    \item Basic requirements of typography are fulfilled: Forbidden
%      linebreaks, correct spacing and the numerator automatically
%      aligning with the height of an ``M''.
%    \item Easy configuration by use of package options.
%    \end{itemize}
%    \section{Bugs again \winkey}
%    Ok, ok, you got me.
%    \begin{itemize}
%    \item Fonts without ``M'' do not work correctly.  Do you know one?
%    \item Fractions in \cmd{\scriptscriptstyle} look ugly because they
%      exceed the height of an ``M''.  As far as I know this is a
%      \LaTeX{} problem, because there are no smaller math fonts
%      available.
%    \item The kerning between numerator and slash or denominator and
%      slash is bad.  In fact, there is none.
%    \item The combination of the \option{ugly} option with text mode
%      can lead to ambiguous fractions.  Be happy that a warning is
%      issued.  Why do you use this option?  Your boss?  Ah, I see.
%    \end{itemize}
%    \section{Implementation}
%    \subsection{Documentation Driver}
%    \begin{macrocode}
%<*driver>
\documentclass[12pt,a4paper]{ltxdoc} 
\usepackage{units}
\begin{document}
  \DocInput{units.dtx}
\end{document}
%</driver>
%    \end{macrocode}
%    \StopEventually{
%      \font\logosl=logosl10 scaled\magstep1
%      \begin{thebibliography}{1}
%      \bibitem{Goossens:1994}
%        \textsc{Goossens, M.; Mittelbach, F.; Samarin, A.}: \emph{Der
%        \LaTeX-Begleiter}.  \newblock Addison-Wesley, Bonn, 1st edn.,
%        1994.
%      \bibitem{Kopka:1994}
%        \textsc{Kopka, H.}: \emph{Einf{\"u}hrung}, vol.~1 of
%        \emph{\LaTeX}.  \newblock Addison-Wesley, Bonn, 1st edn.,
%        1994.
%      \bibitem{Kopka:1995}
%        \textsc{Kopka, H.}: \emph{Erg{\"a}nzungen -- mit einer
%        Einf{\"u}hrung in {\logosl META}\-{\logosl FONT}}, vol.~2 of
%        \emph{\LaTeX}.  \newblock Addison-Wesley, Bonn, 1st edn.,
%        1995.
%      \bibitem{Kopka:1997}
%        \textsc{Kopka, H.}: \emph{Erweiterungen}, vol.~3 of
%        \emph{\LaTeX}.  \newblock Addison-Wesley-Longman, Bonn, 1997.
%      \bibitem{Willberg:1997}
%        \textsc{Willberg, H.\,P.; Forssman, F.}:
%        \emph{Lesetypographie}.  \newblock Schmidt, Mainz, 1997.
%      \end{thebibliography}
%    }
%    \subsection{\file{units.sty}}
%    \begin{macrocode}
%<*units>
%    \end{macrocode}
%    \subsubsection{Identification}
%    As this package uses the \cmd{\RequirePackage} command, it does
%    not work with older \LaTeXe{} versions.
%    \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}[1995/12/01]
%    \end{macrocode}
%    The package identifies itself with its release date, a version
%    number, and a short description.
%    \begin{macrocode}
\ProvidesPackage{units}[1998/08/04 v0.9b Typesetting units]
%    \end{macrocode}
%    \subsubsection{Initialization}
%    The \package{ifthen} package is loaded because a new boolean
%    variable has to be declared.
%    \begin{macrocode}
\RequirePackage{ifthen}
\newboolean{B@UnitsLoose}
%    \end{macrocode}
%    \subsubsection{Option Declaration}
%    This boolean variable is set according to the package options.
%    \begin{macrocode}
\DeclareOption{tight}{\setboolean{B@UnitsLoose}{false}}
\DeclareOption{loose}{\setboolean{B@UnitsLoose}{true}}
%    \end{macrocode}
%    Other options will be passed to the \package{nicefrac} package.
%    \begin{macrocode}
\DeclareOption*{%
  \PassOptionsToPackage{\CurrentOption}{nicefrac}%
}
%    \end{macrocode}
%    \subsubsection{Option Processing}
%    If no options are specified, tight spacing between the value and
%    the dimension of a unit is used by default.  Otherwise the options
%    are processed in the order given by the calling command.
%    \begin{macrocode}
\ExecuteOptions{tight}
\ProcessOptions*
%    \end{macrocode}
%    \subsubsection{Loading Files}
%    \enlargethispage{\baselineskip}
%    Because the macro used for fractions of units is also helpful for
%    other nice fractions, it is build into a stand-alone package
%    called \package{nicefrac}, which therefore is required by
%    \file{units.sty}.
%    \begin{macrocode}
\RequirePackage{nicefrac}[1998/08/04]
%    \end{macrocode}
%    \subsubsection{Defining Commands}
%    In order to make debugging easier for the package user (missing
%    braces or brackets), all commands are declared with the
%    \texttt{*}-form of \cmd{\DeclareRobustCommand}.  The \texttt{*}-form
%    does not allow ``long'' arguments for the commands, so e.\,g.\
%    |\par| commands are forbidden inside the arguments.  This should
%    be no serious restriction.  \winkey
%    \begin{macro}{\unit}
%    First, it is checked if the optional argument is
%    empty.\footnote{Credits go to \person{Bernd Raichle} for
%    supplying this method.}
%    \begin{macrocode}
\DeclareRobustCommand*{\unit}[2][]{%
  \begingroup
    \def\0{#1}%
    \expandafter
  \endgroup
%    \end{macrocode}
%    If it is, nothing happens.
%    \begin{macrocode}
  \ifx\0\@empty
%    \end{macrocode}
%    Otherwise it is typeset and followed by a space depending on the
%    boolean declared above.
%    \begin{macrocode}
  \else
    #1%
    \ifthenelse{\boolean{B@UnitsLoose}}{~}{\,}%
  \fi
%    \end{macrocode}
%    The use of upright fonts for the dimension is forced in math
%    mode.  Units in text mode are typeset with the currently active font.
%    \begin{macrocode}
  \ifthenelse{\boolean{mmode}}{\mathrm{#2}}{#2}%
}
%    \end{macrocode}
%    \end{macro}
%    \begin{macro}{\unitfrac}
%    Fractions of units are typeset by means of the \cmd{\nicefrac}
%    command.  Upright fonts are forced in math mode by use of the
%    optional argument.  The rest of code is identical to the former
%    macro.
%    \begin{macrocode}
\DeclareRobustCommand*{\unitfrac}[3][]{%
  \begingroup
    \def\0{#1}%
    \expandafter
  \endgroup
  \ifx\0\@empty
  \else
    #1%
    \ifthenelse{\boolean{B@UnitsLoose}}{~}{\,}%
  \fi
  \ifthenelse{\boolean{mmode}}{%
    \nicefrac[\mathrm]{#2}{#3}%
  }%
  {%
    \nicefrac{#2}{#3}%
  }%
}
%</units>
%    \end{macrocode}
%    \end{macro}
%    \subsection{\file{nicefrac.sty}}
%    \begin{macrocode}
%<*nicefrac>
%    \end{macrocode}
%    \subsubsection{Identification}
%    \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}[1995/12/01]
\ProvidesPackage{nicefrac}[1998/08/04 v0.9b Nice fractions]
%    \end{macrocode}
%    \subsubsection{Initialization}
%    First, some new lengths are allocated.
%    \begin{macrocode}
\newlength{\L@UnitsRaiseDisplaystyle}
\newlength{\L@UnitsRaiseTextstyle}
\newlength{\L@UnitsRaiseScriptstyle}
%    \end{macrocode}
%    \subsubsection{Loading Files}
%    The \package{ifthen} package is loaded for easier handling of decisions.
%    \begin{macrocode}
\RequirePackage{ifthen}
%    \end{macrocode}
%    \subsubsection{Defining Commands}
%    \begin{macro}{\@UnitsNiceFrac}
%    Now the code used for nice fractions in math mode:
%    \begin{macrocode}
\DeclareRobustCommand*{\@UnitsNiceFrac}[3][]{%
  \ifthenelse{\boolean{mmode}}{%
%    \end{macrocode}
%    For each of the four different math styles the ``M'' height is
%    measured, taking specified font changing commands into account.
%    \begin{macrocode}
    \settoheight{\L@UnitsRaiseDisplaystyle}{%
      \ensuremath{\displaystyle#1{M}}%
    }%
    \settoheight{\L@UnitsRaiseTextstyle}{%
      \ensuremath{\textstyle#1{M}}%
    }%
    \settoheight{\L@UnitsRaiseScriptstyle}{%
      \ensuremath{\scriptstyle#1{M}}%
    }%
    \settoheight{\@tempdima}{%
      \ensuremath{\scriptscriptstyle#1{M}}%
    }%
%    \end{macrocode}
%    The raise height is calculated by taking the difference between
%    the height of a normal ``M'' and the height of an ``M'' set in
%    the correct numerator size.
%    \begin{macrocode}
    \addtolength{\L@UnitsRaiseDisplaystyle}{%
      -\L@UnitsRaiseScriptstyle%
    }%
    \addtolength{\L@UnitsRaiseTextstyle}{%
      -\L@UnitsRaiseScriptstyle%
    }%
    \addtolength{\L@UnitsRaiseScriptstyle}{-\@tempdima}%
%    \end{macrocode}
%    The height of the numerator above the baseline is dependent on
%    the actual fontsize within the math environment.\footnote{Credits
%    go to \person{Wolfgang May} for telling me of the
%    \cmd{\mathchoice} command.}  The numerator is put into an
%    appropriately raised box, within math mode (the additional
%    \cmd{\ensuremath} is necessary, because the \cmd{\raisebox}
%    command leaves math mode).
%
%    The fontsize for the numerator part is \cmd{\scriptstyle} within
%    \cmd{\displaystyle} and \cmd{\textstyle}, but
%    \cmd{\scriptscriptstyle} within \cmd{\scriptstyle} and
%    \cmd{\scriptscriptstyle} context.
%
%    Then the numerator is typeset either with the upright math font or
%    using the font command specified in the optional argument (if it
%    is given).  Only the math alphabet commands may be used:
%    \cmd{\mathnormal}, \cmd{\mathrm}, \cmd{\mathbf}, \cmd{\mathsf},
%    \cmd{\mathit}, \cmd{\mathtt}, and \cmd{\mathcal}.
%    \begin{macrocode}
    \mathchoice
      {%
        \raisebox{\L@UnitsRaiseDisplaystyle}{%
          \ensuremath{\scriptstyle#1{#2}}%
        }%
      }%
      {%
        \raisebox{\L@UnitsRaiseTextstyle}{%
          \ensuremath{\scriptstyle#1{#2}}%
        }%
      }%
      {%
        \raisebox{\L@UnitsRaiseScriptstyle}{%
          \ensuremath{\scriptscriptstyle#1{#2}}%
        }%
      }%
%    \end{macrocode}
%    Within \cmd{\scriptscriptstyle} context, \LaTeX{} does \emph{not}
%    use even smaller fonts for numerator or denominator.  Following
%    the above calculation scheme would lead to a raise height of
%    \unit[0]{pt}.  Because this could make fractions at least
%    ambiguous (if not wrong, like $\mathrm{J}/\mathrm{kg\cdot K}$ for
%    specific thermal capacity), the numerator in
%    \cmd{\scriptscriptstyle} context is raised by the same amount as
%    in \cmd{\scriptstyle} context and so exceeds the height of an
%    adjacent ``M''.  This is ugly, but cannot be avoided.
%    \begin{macrocode}
      {%
        \raisebox{\L@UnitsRaiseScriptstyle}{%
          \ensuremath{\scriptscriptstyle#1{#2}}%
        }%
      }%
%    \end{macrocode}
%    The slash is typeset with reduced spacing to the numerator and to
%    the denominator.
%    \begin{macrocode}
    \mkern-2mu/\mkern-1mu%
%    \end{macrocode}
%    \changes{v0.9b}{1998-08-04}{\cmd{\bgroup} and \cmd{\egroup}
%    instead of \cmd{\begingroup} and \cmd{\endgroup}}
%    The denominator needs a seperate group to delimit the range of the
%    font commands.\footnote{Credits go to \person{Michael Switkes}
%    for the bug report and to \person{Heiko Oberdiek} for telling me
%    why the fix works.}  The choice for the fontsize and the font
%    changing are made just like for the numerator.
%    \begin{macrocode}
    \bgroup
      \mathchoice
        {\scriptstyle}%
        {\scriptstyle}%
        {\scriptscriptstyle}%
        {\scriptscriptstyle}%
      #1{#3}%
    \egroup
  }%
%    \end{macrocode}
%    Now the code for nice fractions in text mode:
%    \begin{macrocode}
  {%
%    \end{macrocode}
%    First, the height of a normal ``M'' is measured.
%    \begin{macrocode}
    \settoheight{\L@UnitsRaiseTextstyle}{#1{M}}%
%    \end{macrocode}
%    \LaTeX{} offers no commands for relative fontsizes.  To make the
%    macro work for example within footnotes or chapter headings, it
%    is necessary to avoid absolute sizing commands like
%    \cmd{\normalsize} or \cmd{\scriptsize}.  To obtain a fontsize
%    that suits for the numerator and denominator, some code was taken
%    from the \LaTeX{} kernel (\file{latex.ltx}, definition of
%    \cmd{\textsuperscript}).
%
%    In all these measurements the font changing effect of the
%    optional argument is taken into account.  Only the following font
%    commands may be used: \cmd{\textnormal}, \cmd{\textrm},
%    \cmd{\textsf}, \cmd{\texttt}, \cmd{\textbf}, \cmd{\textmd},
%    \cmd{\textup}, \cmd{\textit}, \cmd{\textsc}, and \cmd{\textsl}.
%    \begin{macrocode}
    \settoheight{\@tempdima}{%
      \ensuremath{%
        \mbox{\fontsize\sf@size\z@\selectfont#1{M}}%
      }%
    }%
%    \end{macrocode}
%    Both lengths are subtracted.  The result is used for the raise of
%    the numerator.
%    \begin{macrocode}
    \addtolength{\L@UnitsRaiseTextstyle}{-\@tempdima}%
%    \end{macrocode}
%    Finally, the numerator is typeset in ``relative scriptsize''
%    either with the normal text font or using the font command
%    specified in the optional argument (if it is given).
%    \begin{macrocode}
    \raisebox{\L@UnitsRaiseTextstyle}{%
      \ensuremath{%
        \mbox{\fontsize\sf@size\z@\selectfont#1{#2}}%
      }%
    }%
%    \end{macrocode}
%    The code for the slash and the denominator should now be straight
%    forward.
%    \begin{macrocode}
    \ensuremath{\mkern-2mu}/\ensuremath{\mkern-1mu}%
    \ensuremath{%
      \mbox{\fontsize\sf@size\z@\selectfont#1{#3}}%
    }%
  }%
}
%    \end{macrocode}
%    \end{macro}
%    \clearpage
%    \begin{macro}{\@UnitsUglyFrac}
%    Ugly fractions in math mode~\ldots
%    \begin{macrocode}
\DeclareRobustCommand*{\@UnitsUglyFrac}[3][]{%
  \ifthenelse{\boolean{mmode}}{%
%    \end{macrocode}
%    \ldots~are typeset just like normal fractions.
%    \begin{macrocode}
    \frac{#1{#2}}{#1{#3}}%
  }%
%    \end{macrocode}
%    In text mode~\ldots
%    \begin{macrocode}
  {%
%    \end{macrocode}
%    ~\ldots they are typeset like J/kg\textperiodcentered K.
%    \begin{macrocode}
    #1{#2}/#1{#3}%
%    \end{macrocode}
%    Because the fraction can get ambiguous or even wrong, a warning
%    message is issued.
%    \begin{macrocode}
    \PackageWarning{nicefrac}{%
      You used \protect\nicefrac\space or
      \protect\unitfrac\space in text mode\MessageBreak
      and specified the ``ugly'' option.\MessageBreak
      The fraction may be ambiguous or wrong.\MessageBreak
      Please make sure the denominator is
      correct.\MessageBreak
      If it is, you can safely ignore\MessageBreak
      this warning
    }%
  }%
}
%    \end{macrocode}
%    \end{macro}
%    \subsubsection{Option Declaration}
%    Dependent on the package options, the \cmd{\nicefrac} command
%    typesets either nice or ugly fractions by means of the
%    corresponding macros.
%    \begin{macrocode}
\DeclareOption{nice}{%
  \DeclareRobustCommand*{\nicefrac}{\@UnitsNiceFrac}%
}
\DeclareOption{ugly}{%
  \DeclareRobustCommand*{\nicefrac}{\@UnitsUglyFrac}%
}
%    \end{macrocode}
%    \subsubsection{Option Processing}
%    If no options are specified, nice fractions are used by default.
%    Otherwise the options are processed in the order given by the
%    calling command.
%    \begin{macrocode}
\ExecuteOptions{nice}
\ProcessOptions*
%</nicefrac>
%    \end{macrocode}
%    \Finale
\endinput