Debian Codegraph @

texmf-dist/source/latex/units/units.dtx

raw
1 % \iffalse
2 %
3 % units.dtx
4 % Docstrip archive, run through LaTeX.
5 %
6 % Copyright (C) 1998 Axel Reichert
7 % See the files README and COPYING.
8 %
9 % \fi
10 %
11 % \CheckSum{174}
12 %
13 % \changes{v0.9a}{1998-06-25}{First release}
14 % \changes{v0.9b}{1998-08-04}{Documentation update}
15 %
16 %% \CharacterTable
17 %% {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
18 %% 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
19 %% Digits \0\1\2\3\4\5\6\7\8\9
20 %% Exclamation \! Double quote \" Hash (number) \#
21 %% Dollar \$ Percent \% Ampersand \&
22 %% Acute accent \' Left paren \( Right paren \)
23 %% Asterisk \* Plus \+ Comma \,
24 %% Minus \- Point \. Solidus \/
25 %% Colon \: Semicolon \; Less than \<
26 %% Equals \= Greater than \> Question mark \?
27 %% Commercial at \@ Left bracket \[ Backslash \\
28 %% Right bracket \] Circumflex \^ Underscore \_
29 %% Grave accent \` Left brace \{ Vertical bar \|
30 %% Right brace \} Tilde \~}
31 %
32 % \newcommand{\file}[1]{\texttt{#1}}
33 % \newcommand{\option}[1]{\texttt{#1}}
34 % \newcommand{\package}[1]{\texttt{#1}}
35 % \newcommand{\person}[1]{\textsc{#1}}
36 % \newcommand{\winkey}{\texttt{(-;}}
37 %
38 % \title{\file{units.sty} -- \file{nicefrac.sty}}
39 % \author{Axel Reichert}
40 % \date{1998-08-04}
41 % \maketitle
42 % \begin{abstract}
43 % \file{units.sty} is a package for setting units in a
44 % typographically correct way. It is based upon
45 % \file{nicefrac.sty}, a package for nice fractions. See the
46 % files \file{README} and \file{COPYING} for additional
47 % information.
48 % \end{abstract}
49 % \section{Loading}
50 % \begin{tabbing}
51 % Only nice fractions: \= |\usepackage{nicefrac}| \\
52 % Only units or both: \> |\usepackage{units}|
53 % \end{tabbing}
54 % \section{Options}
55 % \begin{tabbing}
56 % Tight spacing for units (default): \=
57 % |\usepackage[tight]{units}| \\
58 % Loose spacing for units: \>
59 % |\usepackage[loose]{units}| \\
60 % ``Nice'' fractions (default): \>
61 % |\usepackage[nice]{nicefrac}| \\
62 % ``Ugly'' fractions: \>
63 % |\usepackage[ugly]{nicefrac}|
64 % \end{tabbing}
65 %
66 % The options \option{nice} and \option{ugly} can also be used for
67 % the \package{units} package, they will simply be passed to the
68 % \package{nicefrac} package, so you can combine the options,
69 % e.\,g.: |\usepackage[loose,ugly]{units}|
70 %
71 % Tight spacing means \cmd{\,} for the space between the value and the
72 % dimension, loose spacing uses |~|, like \unit[1]{m} and 1~m. Nice
73 % fractions look like \nicefrac{m}{s}, ugly fractions look like m/s
74 % in text mode and $\frac{\mathrm{m}}{\mathrm{s}}$ in math mode.
75 % \section{Commands}
76 % \begin{tabbing}
77 % Fractions of units: \=
78 % |\unitfrac[|\meta{val}|]{|\meta{num}|}{|\meta{denom}|}| \kill
79 % Units: \> |\unit[|\meta{val}|]{|\meta{dim}|}| \\
80 % Fractions of units: \>
81 % |\unitfrac[|\meta{val}|]{|\meta{num}|}{|\meta{denom}|}| \\
82 % Nice fractions: \>
83 % |\nicefrac[|\meta{fontcmd}|]{|\meta{num}|}{|\meta{denom}|}|
84 % \end{tabbing}
85 % In these list, \meta{val} and \meta{dim} denote the value and the
86 % dimension of the unit, respectively. \meta{num} and \meta{denom}
87 % are the numerator and denominator of the fraction, and
88 % \meta{fontcmd} can be an author command for fonts or a math
89 % alphabet, see \file{fntguide.dvi}.
90 %
91 % Typically, \meta{val} is only a number and \meta{num} and
92 % \meta{denom} are relatively simple \LaTeX{} expressions. If you
93 % really feel the need for putting whole paragraphs or complex
94 % formula into such a fraction, you have misunderstood the purpose
95 % of this package. \winkey
96 %
97 % A common mistake is to forget the enclosing braces when using the
98 % optional argument of the commands inside another optional
99 % argument. This is no package bug, but a \LaTeX{} feature. Correct
100 % example: |\section[{\unit[1]{m}}]{...}|.\footnote{Credits go to
101 % \person{Frank Knappe} for suggesting this documentation update.}
102 %
103 % It is very important to be aware of the fact that all these
104 % commands distinguish between text mode and math mode. Within text
105 % mode, the font of the surrounding text will be used by default,
106 % for math mode \cmd{\mathrm} is the default.
107 %
108 % This is quite sensible, because you would not want your
109 % collection of delicious recipes (typeset in a mega-cool ultra
110 % condensed bold italic calligraphical font) contain those spindle
111 % Computer Modern Roman just for half a litre of milk. \winkey
112 %
113 % Otherwise, when working on scientific papers, strict and
114 % consistent notation is really a virtue, and, like it or not,
115 % units are typeset with upright fonts. So take great care when
116 % deciding about math mode or not.
117 % \section{Examples}
118 % \begin{tabbing}
119 % |\sffamily\itshape$\unitfrac[1]{m}{s}$| \quad \=
120 % \sffamily\itshape$\unitfrac[1]{m}{s}$ \kill
121 % |\sffamily\bfseries\unit{m}| \> \sffamily\bfseries\unit{m} \\
122 % |\sffamily\bfseries$\unit{m}$| \> \sffamily\bfseries$\unit{m}$ \\
123 % |\sffamily\itshape\unit[1]{m}| \> \sffamily\itshape\unit[1]{m} \\
124 % |\sffamily\itshape$\unit[1]{m}$| \> \sffamily\itshape$\unit[1]{m}$ \\
125 % |\sffamily\bfseries\unitfrac{m}{s}| \>
126 % \sffamily\bfseries\unitfrac{m}{s} \\
127 % |\sffamily\bfseries$\unitfrac{m}{s}$| \>
128 % \sffamily\bfseries$\unitfrac{m}{s}$ \\
129 % |\sffamily\itshape\unitfrac[1]{m}{s}| \>
130 % \sffamily\itshape\unitfrac[1]{m}{s} \\
131 % |\sffamily\itshape$\unitfrac[1]{m}{s}$| \>
132 % \sffamily\itshape$\unitfrac[1]{m}{s}$
133 % \end{tabbing}
134 % As you can see, font changes are ignored in math mode~\ldots
135 % \begin{tabbing}
136 % |\scriptsize\sffamily\itshape$\unitfrac[1]{m}{s}$| \quad \=
137 % \scriptsize\sffamily\itshape$\unitfrac[1]{m}{s}$ \kill
138 % |\scriptsize\sffamily\itshape\unitfrac[1]{m}{s}| \>
139 % \scriptsize\sffamily\itshape\unitfrac[1]{m}{s} \\
140 % |\scriptsize\sffamily\itshape$\unitfrac[1]{m}{s}$| \>
141 % \scriptsize\sffamily\itshape$\unitfrac[1]{m}{s}$
142 % \end{tabbing}
143 % \ldots~except for the fontsize.
144 % \begin{tabbing}
145 % |\bfseries\itshape$\nicefrac{1}{2}$| \quad \=
146 % \bfseries\itshape$\nicefrac{1}{2}$ \kill
147 % |\bfseries\itshape\nicefrac{1}{2}| \>
148 % \bfseries\itshape\nicefrac{1}{2} \\
149 % |\bfseries\itshape$\nicefrac{1}{2}$| \>
150 % \bfseries\itshape$\nicefrac{1}{2}$ \\
151 % |\nicefrac[\texttt]{1}{2}| \> \nicefrac[\texttt]{1}{2} \\
152 % |\nicefrac[\texttt]{\textit{1}}{2}| \>
153 % \nicefrac[\texttt]{\textit{1}}{2} \\
154 % |$\nicefrac[\mathcal]{A}{B}$| \> $\nicefrac[\mathcal]{A}{B}$
155 % \end{tabbing}
156 % The \cmd{\nicefrac} command can deal even with quite strange font
157 % changing commands.
158 % \section{Typography}
159 % Why should units be typeset in upright fonts, not in italics?
160 % Because they have to be distinguished from normal variables:
161 % ``\unit{m}'' is meter, ``$m$'' is a variable, for example a mass.
162 %
163 % Why should the space between the value and the dimension be
164 % non-breakable? Because the reader is disturbed by linebreaks like
165 % 1 $\mathrm{m}$.
166 %
167 % Why should the space between the value and the dimension be a
168 % half word space only? Because things belonging together are
169 % typeset tighter. Compare \unit[1]{m} with 1~m and the normal word
170 % spacing, which can vary from line to line.
171 %
172 % Why should nice fractions be typeset so that the numerator does
173 % not extend above the height of the letter ``M'' and the
174 % denominator does not extend below the baseline? Because otherwise
175 % a stretching of the baselineskip could be necessary due to
176 % descenders.
177 %
178 % By the way, a very common mistake is to place units into
179 % brackets, like [\unit{N}]. The correct notation is $[F] =
180 % \unit{N}$. The brackets indeed are a function with a variable as
181 % an argument. The value returned is the unit. If you need to
182 % specify units in table headings, then use a single row for the
183 % units, where you put them into \emph{parentheses} instead.
184 % \section{Bugs}
185 % None, are you kidding?
186 % \section{Features}
187 % \begin{itemize}
188 % \item Consistent and logical markup of units is enhanced instead
189 % of fiddling around with spacing and fonts.
190 % \item The same command works in text mode \emph{and} math mode,
191 % font and size are adjusted automatically for nice integration
192 % within text mode while forcing strict notation in math mode.
193 % \item Basic requirements of typography are fulfilled: Forbidden
194 % linebreaks, correct spacing and the numerator automatically
195 % aligning with the height of an ``M''.
196 % \item Easy configuration by use of package options.
197 % \end{itemize}
198 % \section{Bugs again \winkey}
199 % Ok, ok, you got me.
200 % \begin{itemize}
201 % \item Fonts without ``M'' do not work correctly. Do you know one?
202 % \item Fractions in \cmd{\scriptscriptstyle} look ugly because they
203 % exceed the height of an ``M''. As far as I know this is a
204 % \LaTeX{} problem, because there are no smaller math fonts
205 % available.
206 % \item The kerning between numerator and slash or denominator and
207 % slash is bad. In fact, there is none.
208 % \item The combination of the \option{ugly} option with text mode
209 % can lead to ambiguous fractions. Be happy that a warning is
210 % issued. Why do you use this option? Your boss? Ah, I see.
211 % \end{itemize}
212 % \section{Implementation}
213 % \subsection{Documentation Driver}
214 % \begin{macrocode}
215 %<*driver>
216 \documentclass[12pt,a4paper]{ltxdoc}
217 \usepackage{units}
218 \begin{document}
219 \DocInput{units.dtx}
220 \end{document}
221 %</driver>
222 % \end{macrocode}
223 % \StopEventually{
224 % \font\logosl=logosl10 scaled\magstep1
225 % \begin{thebibliography}{1}
226 % \bibitem{Goossens:1994}
227 % \textsc{Goossens, M.; Mittelbach, F.; Samarin, A.}: \emph{Der
228 % \LaTeX-Begleiter}. \newblock Addison-Wesley, Bonn, 1st edn.,
229 % 1994.
230 % \bibitem{Kopka:1994}
231 % \textsc{Kopka, H.}: \emph{Einf{\"u}hrung}, vol.~1 of
232 % \emph{\LaTeX}. \newblock Addison-Wesley, Bonn, 1st edn.,
233 % 1994.
234 % \bibitem{Kopka:1995}
235 % \textsc{Kopka, H.}: \emph{Erg{\"a}nzungen -- mit einer
236 % Einf{\"u}hrung in {\logosl META}\-{\logosl FONT}}, vol.~2 of
237 % \emph{\LaTeX}. \newblock Addison-Wesley, Bonn, 1st edn.,
238 % 1995.
239 % \bibitem{Kopka:1997}
240 % \textsc{Kopka, H.}: \emph{Erweiterungen}, vol.~3 of
241 % \emph{\LaTeX}. \newblock Addison-Wesley-Longman, Bonn, 1997.
242 % \bibitem{Willberg:1997}
243 % \textsc{Willberg, H.\,P.; Forssman, F.}:
244 % \emph{Lesetypographie}. \newblock Schmidt, Mainz, 1997.
245 % \end{thebibliography}
246 % }
247 % \subsection{\file{units.sty}}
248 % \begin{macrocode}
249 %<*units>
250 % \end{macrocode}
251 % \subsubsection{Identification}
252 % As this package uses the \cmd{\RequirePackage} command, it does
253 % not work with older \LaTeXe{} versions.
254 % \begin{macrocode}
255 \NeedsTeXFormat{LaTeX2e}[1995/12/01]
256 % \end{macrocode}
257 % The package identifies itself with its release date, a version
258 % number, and a short description.
259 % \begin{macrocode}
260 \ProvidesPackage{units}[1998/08/04 v0.9b Typesetting units]
261 % \end{macrocode}
262 % \subsubsection{Initialization}
263 % The \package{ifthen} package is loaded because a new boolean
264 % variable has to be declared.
265 % \begin{macrocode}
266 \RequirePackage{ifthen}
267 \newboolean{B@UnitsLoose}
268 % \end{macrocode}
269 % \subsubsection{Option Declaration}
270 % This boolean variable is set according to the package options.
271 % \begin{macrocode}
272 \DeclareOption{tight}{\setboolean{B@UnitsLoose}{false}}
273 \DeclareOption{loose}{\setboolean{B@UnitsLoose}{true}}
274 % \end{macrocode}
275 % Other options will be passed to the \package{nicefrac} package.
276 % \begin{macrocode}
277 \DeclareOption*{%
278 \PassOptionsToPackage{\CurrentOption}{nicefrac}%
279 }
280 % \end{macrocode}
281 % \subsubsection{Option Processing}
282 % If no options are specified, tight spacing between the value and
283 % the dimension of a unit is used by default. Otherwise the options
284 % are processed in the order given by the calling command.
285 % \begin{macrocode}
286 \ExecuteOptions{tight}
287 \ProcessOptions*
288 % \end{macrocode}
289 % \subsubsection{Loading Files}
290 % \enlargethispage{\baselineskip}
291 % Because the macro used for fractions of units is also helpful for
292 % other nice fractions, it is build into a stand-alone package
293 % called \package{nicefrac}, which therefore is required by
294 % \file{units.sty}.
295 % \begin{macrocode}
296 \RequirePackage{nicefrac}[1998/08/04]
297 % \end{macrocode}
298 % \subsubsection{Defining Commands}
299 % In order to make debugging easier for the package user (missing
300 % braces or brackets), all commands are declared with the
301 % \texttt{*}-form of \cmd{\DeclareRobustCommand}. The \texttt{*}-form
302 % does not allow ``long'' arguments for the commands, so e.\,g.\
303 % |\par| commands are forbidden inside the arguments. This should
304 % be no serious restriction. \winkey
305 % \begin{macro}{\unit}
306 % First, it is checked if the optional argument is
307 % empty.\footnote{Credits go to \person{Bernd Raichle} for
308 % supplying this method.}
309 % \begin{macrocode}
310 \DeclareRobustCommand*{\unit}[2][]{%
311 \begingroup
312 \def\0{#1}%
313 \expandafter
314 \endgroup
315 % \end{macrocode}
316 % If it is, nothing happens.
317 % \begin{macrocode}
318 \ifx\0\@empty
319 % \end{macrocode}
320 % Otherwise it is typeset and followed by a space depending on the
321 % boolean declared above.
322 % \begin{macrocode}
323 \else
324 #1%
325 \ifthenelse{\boolean{B@UnitsLoose}}{~}{\,}%
326 \fi
327 % \end{macrocode}
328 % The use of upright fonts for the dimension is forced in math
329 % mode. Units in text mode are typeset with the currently active font.
330 % \begin{macrocode}
331 \ifthenelse{\boolean{mmode}}{\mathrm{#2}}{#2}%
332 }
333 % \end{macrocode}
334 % \end{macro}
335 % \begin{macro}{\unitfrac}
336 % Fractions of units are typeset by means of the \cmd{\nicefrac}
337 % command. Upright fonts are forced in math mode by use of the
338 % optional argument. The rest of code is identical to the former
339 % macro.
340 % \begin{macrocode}
341 \DeclareRobustCommand*{\unitfrac}[3][]{%
342 \begingroup
343 \def\0{#1}%
344 \expandafter
345 \endgroup
346 \ifx\0\@empty
347 \else
348 #1%
349 \ifthenelse{\boolean{B@UnitsLoose}}{~}{\,}%
350 \fi
351 \ifthenelse{\boolean{mmode}}{%
352 \nicefrac[\mathrm]{#2}{#3}%
353 }%
354 {%
355 \nicefrac{#2}{#3}%
356 }%
357 }
358 %</units>
359 % \end{macrocode}
360 % \end{macro}
361 % \subsection{\file{nicefrac.sty}}
362 % \begin{macrocode}
363 %<*nicefrac>
364 % \end{macrocode}
365 % \subsubsection{Identification}
366 % \begin{macrocode}
367 \NeedsTeXFormat{LaTeX2e}[1995/12/01]
368 \ProvidesPackage{nicefrac}[1998/08/04 v0.9b Nice fractions]
369 % \end{macrocode}
370 % \subsubsection{Initialization}
371 % First, some new lengths are allocated.
372 % \begin{macrocode}
373 \newlength{\L@UnitsRaiseDisplaystyle}
374 \newlength{\L@UnitsRaiseTextstyle}
375 \newlength{\L@UnitsRaiseScriptstyle}
376 % \end{macrocode}
377 % \subsubsection{Loading Files}
378 % The \package{ifthen} package is loaded for easier handling of decisions.
379 % \begin{macrocode}
380 \RequirePackage{ifthen}
381 % \end{macrocode}
382 % \subsubsection{Defining Commands}
383 % \begin{macro}{\@UnitsNiceFrac}
384 % Now the code used for nice fractions in math mode:
385 % \begin{macrocode}
386 \DeclareRobustCommand*{\@UnitsNiceFrac}[3][]{%
387 \ifthenelse{\boolean{mmode}}{%
388 % \end{macrocode}
389 % For each of the four different math styles the ``M'' height is
390 % measured, taking specified font changing commands into account.
391 % \begin{macrocode}
392 \settoheight{\L@UnitsRaiseDisplaystyle}{%
393 \ensuremath{\displaystyle#1{M}}%
394 }%
395 \settoheight{\L@UnitsRaiseTextstyle}{%
396 \ensuremath{\textstyle#1{M}}%
397 }%
398 \settoheight{\L@UnitsRaiseScriptstyle}{%
399 \ensuremath{\scriptstyle#1{M}}%
400 }%
401 \settoheight{\@tempdima}{%
402 \ensuremath{\scriptscriptstyle#1{M}}%
403 }%
404 % \end{macrocode}
405 % The raise height is calculated by taking the difference between
406 % the height of a normal ``M'' and the height of an ``M'' set in
407 % the correct numerator size.
408 % \begin{macrocode}
409 \addtolength{\L@UnitsRaiseDisplaystyle}{%
410 -\L@UnitsRaiseScriptstyle%
411 }%
412 \addtolength{\L@UnitsRaiseTextstyle}{%
413 -\L@UnitsRaiseScriptstyle%
414 }%
415 \addtolength{\L@UnitsRaiseScriptstyle}{-\@tempdima}%
416 % \end{macrocode}
417 % The height of the numerator above the baseline is dependent on
418 % the actual fontsize within the math environment.\footnote{Credits
419 % go to \person{Wolfgang May} for telling me of the
420 % \cmd{\mathchoice} command.} The numerator is put into an
421 % appropriately raised box, within math mode (the additional
422 % \cmd{\ensuremath} is necessary, because the \cmd{\raisebox}
423 % command leaves math mode).
424 %
425 % The fontsize for the numerator part is \cmd{\scriptstyle} within
426 % \cmd{\displaystyle} and \cmd{\textstyle}, but
427 % \cmd{\scriptscriptstyle} within \cmd{\scriptstyle} and
428 % \cmd{\scriptscriptstyle} context.
429 %
430 % Then the numerator is typeset either with the upright math font or
431 % using the font command specified in the optional argument (if it
432 % is given). Only the math alphabet commands may be used:
433 % \cmd{\mathnormal}, \cmd{\mathrm}, \cmd{\mathbf}, \cmd{\mathsf},
434 % \cmd{\mathit}, \cmd{\mathtt}, and \cmd{\mathcal}.
435 % \begin{macrocode}
436 \mathchoice
437 {%
438 \raisebox{\L@UnitsRaiseDisplaystyle}{%
439 \ensuremath{\scriptstyle#1{#2}}%
440 }%
441 }%
442 {%
443 \raisebox{\L@UnitsRaiseTextstyle}{%
444 \ensuremath{\scriptstyle#1{#2}}%
445 }%
446 }%
447 {%
448 \raisebox{\L@UnitsRaiseScriptstyle}{%
449 \ensuremath{\scriptscriptstyle#1{#2}}%
450 }%
451 }%
452 % \end{macrocode}
453 % Within \cmd{\scriptscriptstyle} context, \LaTeX{} does \emph{not}
454 % use even smaller fonts for numerator or denominator. Following
455 % the above calculation scheme would lead to a raise height of
456 % \unit[0]{pt}. Because this could make fractions at least
457 % ambiguous (if not wrong, like $\mathrm{J}/\mathrm{kg\cdot K}$ for
458 % specific thermal capacity), the numerator in
459 % \cmd{\scriptscriptstyle} context is raised by the same amount as
460 % in \cmd{\scriptstyle} context and so exceeds the height of an
461 % adjacent ``M''. This is ugly, but cannot be avoided.
462 % \begin{macrocode}
463 {%
464 \raisebox{\L@UnitsRaiseScriptstyle}{%
465 \ensuremath{\scriptscriptstyle#1{#2}}%
466 }%
467 }%
468 % \end{macrocode}
469 % The slash is typeset with reduced spacing to the numerator and to
470 % the denominator.
471 % \begin{macrocode}
472 \mkern-2mu/\mkern-1mu%
473 % \end{macrocode}
474 % \changes{v0.9b}{1998-08-04}{\cmd{\bgroup} and \cmd{\egroup}
475 % instead of \cmd{\begingroup} and \cmd{\endgroup}}
476 % The denominator needs a seperate group to delimit the range of the
477 % font commands.\footnote{Credits go to \person{Michael Switkes}
478 % for the bug report and to \person{Heiko Oberdiek} for telling me
479 % why the fix works.} The choice for the fontsize and the font
480 % changing are made just like for the numerator.
481 % \begin{macrocode}
482 \bgroup
483 \mathchoice
484 {\scriptstyle}%
485 {\scriptstyle}%
486 {\scriptscriptstyle}%
487 {\scriptscriptstyle}%
488 #1{#3}%
489 \egroup
490 }%
491 % \end{macrocode}
492 % Now the code for nice fractions in text mode:
493 % \begin{macrocode}
494 {%
495 % \end{macrocode}
496 % First, the height of a normal ``M'' is measured.
497 % \begin{macrocode}
498 \settoheight{\L@UnitsRaiseTextstyle}{#1{M}}%
499 % \end{macrocode}
500 % \LaTeX{} offers no commands for relative fontsizes. To make the
501 % macro work for example within footnotes or chapter headings, it
502 % is necessary to avoid absolute sizing commands like
503 % \cmd{\normalsize} or \cmd{\scriptsize}. To obtain a fontsize
504 % that suits for the numerator and denominator, some code was taken
505 % from the \LaTeX{} kernel (\file{latex.ltx}, definition of
506 % \cmd{\textsuperscript}).
507 %
508 % In all these measurements the font changing effect of the
509 % optional argument is taken into account. Only the following font
510 % commands may be used: \cmd{\textnormal}, \cmd{\textrm},
511 % \cmd{\textsf}, \cmd{\texttt}, \cmd{\textbf}, \cmd{\textmd},
512 % \cmd{\textup}, \cmd{\textit}, \cmd{\textsc}, and \cmd{\textsl}.
513 % \begin{macrocode}
514 \settoheight{\@tempdima}{%
515 \ensuremath{%
516 \mbox{\fontsize\sf@size\z@\selectfont#1{M}}%
517 }%
518 }%
519 % \end{macrocode}
520 % Both lengths are subtracted. The result is used for the raise of
521 % the numerator.
522 % \begin{macrocode}
523 \addtolength{\L@UnitsRaiseTextstyle}{-\@tempdima}%
524 % \end{macrocode}
525 % Finally, the numerator is typeset in ``relative scriptsize''
526 % either with the normal text font or using the font command
527 % specified in the optional argument (if it is given).
528 % \begin{macrocode}
529 \raisebox{\L@UnitsRaiseTextstyle}{%
530 \ensuremath{%
531 \mbox{\fontsize\sf@size\z@\selectfont#1{#2}}%
532 }%
533 }%
534 % \end{macrocode}
535 % The code for the slash and the denominator should now be straight
536 % forward.
537 % \begin{macrocode}
538 \ensuremath{\mkern-2mu}/\ensuremath{\mkern-1mu}%
539 \ensuremath{%
540 \mbox{\fontsize\sf@size\z@\selectfont#1{#3}}%
541 }%
542 }%
543 }
544 % \end{macrocode}
545 % \end{macro}
546 % \clearpage
547 % \begin{macro}{\@UnitsUglyFrac}
548 % Ugly fractions in math mode~\ldots
549 % \begin{macrocode}
550 \DeclareRobustCommand*{\@UnitsUglyFrac}[3][]{%
551 \ifthenelse{\boolean{mmode}}{%
552 % \end{macrocode}
553 % \ldots~are typeset just like normal fractions.
554 % \begin{macrocode}
555 \frac{#1{#2}}{#1{#3}}%
556 }%
557 % \end{macrocode}
558 % In text mode~\ldots
559 % \begin{macrocode}
560 {%
561 % \end{macrocode}
562 % ~\ldots they are typeset like J/kg\textperiodcentered K.
563 % \begin{macrocode}
564 #1{#2}/#1{#3}%
565 % \end{macrocode}
566 % Because the fraction can get ambiguous or even wrong, a warning
567 % message is issued.
568 % \begin{macrocode}
569 \PackageWarning{nicefrac}{%
570 You used \protect\nicefrac\space or
571 \protect\unitfrac\space in text mode\MessageBreak
572 and specified the ``ugly'' option.\MessageBreak
573 The fraction may be ambiguous or wrong.\MessageBreak
574 Please make sure the denominator is
575 correct.\MessageBreak
576 If it is, you can safely ignore\MessageBreak
577 this warning
578 }%
579 }%
580 }
581 % \end{macrocode}
582 % \end{macro}
583 % \subsubsection{Option Declaration}
584 % Dependent on the package options, the \cmd{\nicefrac} command
585 % typesets either nice or ugly fractions by means of the
586 % corresponding macros.
587 % \begin{macrocode}
588 \DeclareOption{nice}{%
589 \DeclareRobustCommand*{\nicefrac}{\@UnitsNiceFrac}%
590 }
591 \DeclareOption{ugly}{%
592 \DeclareRobustCommand*{\nicefrac}{\@UnitsUglyFrac}%
593 }
594 % \end{macrocode}
595 % \subsubsection{Option Processing}
596 % If no options are specified, nice fractions are used by default.
597 % Otherwise the options are processed in the order given by the
598 % calling command.
599 % \begin{macrocode}
600 \ExecuteOptions{nice}
601 \ProcessOptions*
602 %</nicefrac>
603 % \end{macrocode}
604 % \Finale
605 \endinput