| 1 |
% \iffalse meta-comment |
| 2 |
% |
| 3 |
% Copyright (C) 2016-2017 by Takahiro Ueda <[email protected]> |
| 4 |
% |
| 5 |
% This file may be distributed and/or modified under the conditions of |
| 6 |
% the LaTeX Project Public License, either version 1.3 of this license |
| 7 |
% or (at your option) any later version. The latest version of this |
| 8 |
% license is in: |
| 9 |
% |
| 10 |
% http://www.latex-project.org/lppl.txt |
| 11 |
% |
| 12 |
% and version 1.3 or later is part of all distributions of LaTeX |
| 13 |
% version 2005/12/01 or later. |
| 14 |
% |
| 15 |
% To typeset this document, type the following commands: |
| 16 |
% latex autobreak.dtx |
| 17 |
% makeindex -s gind.ist autobreak.idx |
| 18 |
% makeindex -s gglo.ist -o autobreak.gls autobreak.glo |
| 19 |
% latex autobreak.dtx |
| 20 |
% latex autobreak.dtx |
| 21 |
% |
| 22 |
%<*ignore> |
| 23 |
\def\nameofplainTeX{plain} |
| 24 |
\ifx\fmtname\nameofplainTeX\else |
| 25 |
\expandafter\begingroup |
| 26 |
\fi |
| 27 |
%</ignore> |
| 28 |
%<*install> |
| 29 |
|
| 30 |
\input docstrip.tex |
| 31 |
\keepsilent |
| 32 |
\askforoverwritefalse |
| 33 |
\preamble |
| 34 |
|
| 35 |
This is a generated file. |
| 36 |
|
| 37 |
Copyright (C) 2016-2017 by Takahiro Ueda <[email protected]> |
| 38 |
|
| 39 |
This file may be distributed and/or modified under the conditions of |
| 40 |
the LaTeX Project Public License, either version 1.3 of this license |
| 41 |
or (at your option) any later version. The latest version of this |
| 42 |
license is in: |
| 43 |
|
| 44 |
http://www.latex-project.org/lppl.txt |
| 45 |
|
| 46 |
and version 1.3 or later is part of all distributions of LaTeX |
| 47 |
version 2005/12/01 or later. |
| 48 |
|
| 49 |
\endpreamble |
| 50 |
|
| 51 |
\generate{ |
| 52 |
\usedir{tex/latex/autobreak} |
| 53 |
\file{\jobname.sty}{\from{\jobname.dtx}{package}} |
| 54 |
} |
| 55 |
|
| 56 |
%</install> |
| 57 |
%<*ignore> |
| 58 |
\iffalse |
| 59 |
%</ignore> |
| 60 |
%<*install> |
| 61 |
|
| 62 |
\obeyspaces |
| 63 |
\Msg{*************************************************************} |
| 64 |
\Msg{* *} |
| 65 |
\Msg{* To finish the installation you have to move the following *} |
| 66 |
\Msg{* file into a directory searched by TeX: *} |
| 67 |
\Msg{* *} |
| 68 |
\Msg{* autobreak.sty *} |
| 69 |
\Msg{* *} |
| 70 |
\Msg{* To produce the documentation run the file autobreak.dtx *} |
| 71 |
\Msg{* through LaTeX. *} |
| 72 |
\Msg{* *} |
| 73 |
\Msg{* Happy TeXing! *} |
| 74 |
\Msg{* *} |
| 75 |
\Msg{*************************************************************} |
| 76 |
|
| 77 |
\endbatchfile |
| 78 |
|
| 79 |
%</install> |
| 80 |
%<*ignore> |
| 81 |
\fi |
| 82 |
\generate{ |
| 83 |
\file{\jobname.ins}{\from{\jobname.dtx}{install}} |
| 84 |
} |
| 85 |
\ifx\fmtname\nameofplainTeX |
| 86 |
\expandafter\endbatchfile |
| 87 |
\else |
| 88 |
\expandafter\endgroup |
| 89 |
\fi |
| 90 |
%</ignore> |
| 91 |
%<*driver> |
| 92 |
\documentclass{ltxdoc} |
| 93 |
\usepackage[T1]{fontenc} |
| 94 |
\usepackage{lmodern} |
| 95 |
\usepackage{color} |
| 96 |
\usepackage{hypdoc} |
| 97 |
\usepackage{amsmath} |
| 98 |
\usepackage{autobreak} |
| 99 |
\EnableCrossrefs |
| 100 |
\CodelineIndex |
| 101 |
\RecordChanges |
| 102 |
|
| 103 |
\allowdisplaybreaks |
| 104 |
|
| 105 |
% The default value of IndexColumns=3 never fits in pages. |
| 106 |
\setcounter{IndexColumns}{2} |
| 107 |
|
| 108 |
\definecolor{hilite}{rgb}{0.2,0.4,0.7} |
| 109 |
\makeatletter |
| 110 |
\renewcommand{\HDorg@theCodelineNo}{% |
| 111 |
\textcolor{hilite}{\rmfamily\scriptsize\arabic{CodelineNo}}} |
| 112 |
\makeatother |
| 113 |
|
| 114 |
\providecommand\env[1]{\texttt{#1}} |
| 115 |
\providecommand\pkg[1]{\textsf{#1}} |
| 116 |
|
| 117 |
\usepackage{fancyvrb} |
| 118 |
|
| 119 |
\edef\examplefilename{\jobname.tmp} |
| 120 |
|
| 121 |
\newenvironment{example}{% |
| 122 |
\VerbatimEnvironment |
| 123 |
\begin{VerbatimOut}[gobble=2]{\examplefilename}% |
| 124 |
}{% |
| 125 |
\end{VerbatimOut}% |
| 126 |
\smallskip |
| 127 |
\noindent |
| 128 |
\begin{list}{}{}\item |
| 129 |
\begin{minipage}[t]{0.45\linewidth}% |
| 130 |
\VerbatimInput{\examplefilename}% |
| 131 |
\end{minipage}% |
| 132 |
\begin{minipage}[t]{0.5\linewidth}% |
| 133 |
\input{\examplefilename}% |
| 134 |
\end{minipage}% |
| 135 |
\end{list}% |
| 136 |
\smallskip |
| 137 |
\noindent |
| 138 |
} |
| 139 |
|
| 140 |
\newenvironment{example*}{% |
| 141 |
\VerbatimEnvironment |
| 142 |
\begin{VerbatimOut}[gobble=2]{\examplefilename}% |
| 143 |
}{% |
| 144 |
\end{VerbatimOut}% |
| 145 |
\smallskip |
| 146 |
\noindent |
| 147 |
\begin{list}{}{}\item |
| 148 |
\begin{minipage}[t]{0.95\linewidth}% |
| 149 |
\VerbatimInput{\examplefilename}% |
| 150 |
\end{minipage}% |
| 151 |
\end{list}% |
| 152 |
\smallskip |
| 153 |
\noindent |
| 154 |
} |
| 155 |
|
| 156 |
\newenvironment{example**}{% |
| 157 |
\VerbatimEnvironment |
| 158 |
\begin{VerbatimOut}[gobble=2]{\examplefilename}% |
| 159 |
}{% |
| 160 |
\end{VerbatimOut}% |
| 161 |
\smallskip |
| 162 |
\noindent |
| 163 |
\begin{list}{}{}\item |
| 164 |
\begin{minipage}[t]{0.95\linewidth}% |
| 165 |
\VerbatimInput{\examplefilename}% |
| 166 |
\end{minipage} \\ |
| 167 |
\begin{minipage}[t]{0.95\linewidth}% |
| 168 |
\input{\examplefilename}% |
| 169 |
\end{minipage}% |
| 170 |
\end{list}% |
| 171 |
\smallskip |
| 172 |
\noindent |
| 173 |
} |
| 174 |
|
| 175 |
% decl from ltxguide[2001/05/28] |
| 176 |
\newenvironment{decl}[1][]% |
| 177 |
{\par\small\addvspace{4.5ex plus 1ex}% |
| 178 |
\vskip -\parskip |
| 179 |
\ifx\relax#1\relax |
| 180 |
\def\@decl@date{}% |
| 181 |
\else |
| 182 |
\def\@decl@date{\NEWfeature{#1}}% |
| 183 |
\fi |
| 184 |
\noindent\hspace{-\leftmargini}% |
| 185 |
\begin{tabular}{|l|}\hline\ignorespaces}% |
| 186 |
{\\\hline\end{tabular}\nobreak\@decl@date\par\nobreak |
| 187 |
\vspace{2.3ex}\vskip -\parskip} |
| 188 |
|
| 189 |
\newcommand\NumberedBox[1]{% |
| 190 |
\, \framebox[25pt]{#1} \,% |
| 191 |
} |
| 192 |
|
| 193 |
\newcommand\LongerNumberedBox[1]{% |
| 194 |
\, \framebox[30pt]{#1} \,% |
| 195 |
} |
| 196 |
|
| 197 |
\begin{document} |
| 198 |
\DocInput{\jobname.dtx} |
| 199 |
\end{document} |
| 200 |
%</driver> |
| 201 |
% \fi |
| 202 |
% |
| 203 |
% \CheckSum{409} |
| 204 |
% |
| 205 |
% \CharacterTable |
| 206 |
% {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 |
| 207 |
% 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 |
| 208 |
% Digits \0\1\2\3\4\5\6\7\8\9 |
| 209 |
% Exclamation \! Double quote \" Hash (number) \# |
| 210 |
% Dollar \$ Percent \% Ampersand \& |
| 211 |
% Acute accent \' Left paren \( Right paren \) |
| 212 |
% Asterisk \* Plus \+ Comma \, |
| 213 |
% Minus \- Point \. Solidus \/ |
| 214 |
% Colon \: Semicolon \; Less than \< |
| 215 |
% Equals \= Greater than \> Question mark \? |
| 216 |
% Commercial at \@ Left bracket \[ Backslash \\ |
| 217 |
% Right bracket \] Circumflex \^ Underscore \_ |
| 218 |
% Grave accent \` Left brace \{ Vertical bar \| |
| 219 |
% Right brace \} Tilde \~} |
| 220 |
% |
| 221 |
% \GetFileInfo{\jobname.sty} |
| 222 |
% |
| 223 |
% \title{The \pkg{autobreak} package\thanks{This document corresponds |
| 224 |
% to \pkg{autobreak}~\fileversion, dated \filedate.}} |
| 225 |
% \author{Takahiro Ueda} |
| 226 |
% |
| 227 |
% \date{23 February 2017} |
| 228 |
% |
| 229 |
% \maketitle |
| 230 |
% |
| 231 |
% \begin{abstract} |
| 232 |
% This package implements a simple mechanism of line/page breaking |
| 233 |
% within the \env{align} environment of the \pkg{amsmath} package; new |
| 234 |
% line characters are considered as possible candidates for the breaks |
| 235 |
% and the package tries to put breaks at adequate places. |
| 236 |
% It is suitable for computer-generated long formulae with many terms. |
| 237 |
% \end{abstract} |
| 238 |
% |
| 239 |
% \changes{v0.1}{2016/06/03}{Initial version} |
| 240 |
% |
| 241 |
% \tableofcontents |
| 242 |
% |
| 243 |
% \section{Introduction} |
| 244 |
% |
| 245 |
% Sometimes people want to put long formulae in their documents, which |
| 246 |
% do not fit in a line and may span over multiple pages. |
| 247 |
% The following is an equation of explicitly writing down the first 50 |
| 248 |
% terms in the sum of the well-known Basel problem: |
| 249 |
% \begin{align} |
| 250 |
% \begin{autobreak} |
| 251 |
% \zeta(2) = |
| 252 |
% 1 |
| 253 |
% + \frac{1}{4} |
| 254 |
% + \frac{1}{9} |
| 255 |
% + \frac{1}{16} |
| 256 |
% + \frac{1}{25} |
| 257 |
% + \frac{1}{36} |
| 258 |
% + \frac{1}{49} |
| 259 |
% + \frac{1}{64} |
| 260 |
% + \frac{1}{81} |
| 261 |
% + \frac{1}{100} |
| 262 |
% + \frac{1}{121} |
| 263 |
% + \frac{1}{144} |
| 264 |
% + \frac{1}{169} |
| 265 |
% + \frac{1}{196} |
| 266 |
% + \frac{1}{225} |
| 267 |
% + \frac{1}{256} |
| 268 |
% + \frac{1}{289} |
| 269 |
% + \frac{1}{324} |
| 270 |
% + \frac{1}{361} |
| 271 |
% + \frac{1}{400} |
| 272 |
% + \frac{1}{441} |
| 273 |
% + \frac{1}{484} |
| 274 |
% + \frac{1}{529} |
| 275 |
% + \frac{1}{576} |
| 276 |
% + \frac{1}{625} |
| 277 |
% + \frac{1}{676} |
| 278 |
% + \frac{1}{729} |
| 279 |
% + \frac{1}{784} |
| 280 |
% + \frac{1}{841} |
| 281 |
% + \frac{1}{900} |
| 282 |
% + \frac{1}{961} |
| 283 |
% + \frac{1}{1024} |
| 284 |
% + \frac{1}{1089} |
| 285 |
% + \frac{1}{1156} |
| 286 |
% + \frac{1}{1225} |
| 287 |
% + \frac{1}{1296} |
| 288 |
% + \frac{1}{1369} |
| 289 |
% + \frac{1}{1444} |
| 290 |
% + \frac{1}{1521} |
| 291 |
% + \frac{1}{1600} |
| 292 |
% + \frac{1}{1681} |
| 293 |
% + \frac{1}{1764} |
| 294 |
% + \frac{1}{1849} |
| 295 |
% + \frac{1}{1936} |
| 296 |
% + \frac{1}{2025} |
| 297 |
% + \frac{1}{2116} |
| 298 |
% + \frac{1}{2209} |
| 299 |
% + \frac{1}{2304} |
| 300 |
% + \frac{1}{2401} |
| 301 |
% + \frac{1}{2500} |
| 302 |
% + \dots |
| 303 |
% \end{autobreak} |
| 304 |
% \end{align} |
| 305 |
% The above example might seem nonsense, but putting long formulae may |
| 306 |
% have a meaning in some cases and become inevitable for completeness of |
| 307 |
% documents, writing self-contained papers, or just to impress readers. |
| 308 |
% They are typically generated as outputs of computer algebra systems, |
| 309 |
% and would have the form of a sum of many terms while each term is |
| 310 |
% short. |
| 311 |
% |
| 312 |
% Then, the question is how to break long formulae in such a way that |
| 313 |
% the expressions do not make any overfull lines for \LaTeX{}. |
| 314 |
% Certainly, one can attempt to manually insert line breaks by trial and |
| 315 |
% error, checking whether \LaTeX{} warns overfull lines, and this |
| 316 |
% process could be automatized by external scripts at some extent. |
| 317 |
% A shortcoming of such `manual' approaches is that line breaks have to |
| 318 |
% be reexamined whenever the layout of the document is changed, e.g., |
| 319 |
% replacing the document class or reusing existing equations into |
| 320 |
% another document with a different format. |
| 321 |
% |
| 322 |
% The goal of the \pkg{autobreak} package is to give a reasonably simple |
| 323 |
% solution for (semi-)automatic line breaking of long formulae within |
| 324 |
% \LaTeX{}^^A |
| 325 |
% \footnote{^^A |
| 326 |
% There is another package \pkg{breqn} |
| 327 |
% (\url{https://www.ctan.org/pkg/breqn}), which adopts a more |
| 328 |
% automatic fashion and is useful for more sophisticated line |
| 329 |
% breaking, unless you get ``Dimension too large'' error for really |
| 330 |
% big expressions. |
| 331 |
% }. |
| 332 |
% |
| 333 |
% \section{Usage} |
| 334 |
% |
| 335 |
% The \pkg{autobreak} package is supposed to be used together with the |
| 336 |
% \pkg{amsmath} package^^A |
| 337 |
% \footnote{^^A |
| 338 |
% \url{https://www.ctan.org/pkg/amsmath}. |
| 339 |
% }^^A |
| 340 |
% \footnote{^^A |
| 341 |
% Actually \pkg{autobreak} internally loads \pkg{amsmath}, but it is |
| 342 |
% still a good practice to explicitly include all packages providing |
| 343 |
% macros used in your document. |
| 344 |
% }: |
| 345 |
% \begin{example*} |
| 346 |
% \usepackage{amsmath} |
| 347 |
% \usepackage{autobreak} |
| 348 |
% \end{example*} |
| 349 |
% When your document contains long equations over multiple pages, |
| 350 |
% you might want to use \cs{allowdisplaybreaks} of \pkg{amsmath} |
| 351 |
% package: |
| 352 |
% \begin{example*} |
| 353 |
% \allowdisplaybreaks |
| 354 |
% \end{example*} |
| 355 |
% |
| 356 |
% \begin{decl} |
| 357 |
% |\begin{autobreak}| \\ |
| 358 |
% \meta{long-equations} \\ |
| 359 |
% |\end{autobreak}| |
| 360 |
% \end{decl} |
| 361 |
% \SpecialEnvIndex{autobreak} |
| 362 |
% The \env{autobreak} environment is used for breaking lines in long |
| 363 |
% formulae in the \env{align} environment of \pkg{amsmath}^^A |
| 364 |
% \footnote{^^A |
| 365 |
% Technically, \env{align} (with \cs{notag} to suppress equation |
| 366 |
% numbers except the last line) is the only option we can use for |
| 367 |
% page-break aligned equations within \pkg{amsmath} because |
| 368 |
% \env{split}, \env{gathered}, \env{aligned} and \env{alginedat} do |
| 369 |
% not allow page breaking. |
| 370 |
% \env{dmath} of \env{breqn} with \cs{eqinterlinepenalty}\texttt{=0} |
| 371 |
% allows page breaking, but may fail to find a reasonable tag place. |
| 372 |
% }. |
| 373 |
% \begin{example} |
| 374 |
% \begin{align} |
| 375 |
% \begin{autobreak} |
| 376 |
% \zeta(3) = |
| 377 |
% 1 |
| 378 |
% + \frac{1}{8} |
| 379 |
% + \frac{1}{27} |
| 380 |
% + \frac{1}{64} |
| 381 |
% + \frac{1}{125} |
| 382 |
% + \frac{1}{216} |
| 383 |
% + \frac{1}{343} |
| 384 |
% + \frac{1}{512} |
| 385 |
% + \frac{1}{729} |
| 386 |
% + \frac{1}{1000} |
| 387 |
% + \frac{1}{1331} |
| 388 |
% + \frac{1}{1728} |
| 389 |
% + \frac{1}{2197} |
| 390 |
% + \frac{1}{2744} |
| 391 |
% + \frac{1}{3375} |
| 392 |
% + \frac{1}{4096} |
| 393 |
% + \frac{1}{4913} |
| 394 |
% + \frac{1}{5832} |
| 395 |
% + \frac{1}{6859} |
| 396 |
% + \frac{1}{8000} |
| 397 |
% + \dots |
| 398 |
% \end{autobreak} |
| 399 |
% \end{align} |
| 400 |
% \end{example} |
| 401 |
% The magic happens from the simple fact that \env{autobreak} interprets |
| 402 |
% all new line characters appearing between |\begin{autobreak}| and |
| 403 |
% |\end{autobreak}| as \textit{breakable points}, at which any line |
| 404 |
% breaks can be logically inserted. To be more exact, the first |
| 405 |
% non-empty block, separated from the rest by a new line character, |
| 406 |
% determines the indentation of the successive lines. Then |
| 407 |
% \env{autobreak} tries to fill the line with the rest of the blocks, |
| 408 |
% and puts a line break when they do not fit in a line. This is |
| 409 |
% clarified by the following example: |
| 410 |
% \begin{example} |
| 411 |
% \begin{align} |
| 412 |
% \begin{autobreak} |
| 413 |
% \NumberedBox{1} |
| 414 |
% \NumberedBox{2} |
| 415 |
% \NumberedBox{3} |
| 416 |
% \NumberedBox{4} |
| 417 |
% \NumberedBox{5} |
| 418 |
% \NumberedBox{6} |
| 419 |
% \NumberedBox{7} |
| 420 |
% \NumberedBox{8} |
| 421 |
% \NumberedBox{9} |
| 422 |
% \NumberedBox{10} |
| 423 |
% \NumberedBox{11} |
| 424 |
% \end{autobreak} |
| 425 |
% \end{align} |
| 426 |
% \end{example} |
| 427 |
% |
| 428 |
% It is also possible to put more than one \env{autobreak} in one |
| 429 |
% \env{align}: |
| 430 |
% \begin{example} |
| 431 |
% \begin{align} |
| 432 |
% \begin{autobreak} |
| 433 |
% \NumberedBox{1} = |
| 434 |
% \NumberedBox{2} |
| 435 |
% + \NumberedBox{3} |
| 436 |
% + \NumberedBox{4} |
| 437 |
% + \NumberedBox{5} |
| 438 |
% + \NumberedBox{6} |
| 439 |
% + \NumberedBox{7} |
| 440 |
% + \NumberedBox{8} |
| 441 |
% + \NumberedBox{9} |
| 442 |
% + \NumberedBox{10} , |
| 443 |
% \end{autobreak} |
| 444 |
% \\ |
| 445 |
% \begin{autobreak} |
| 446 |
% \LongerNumberedBox{1} = |
| 447 |
% \NumberedBox{2} |
| 448 |
% + \NumberedBox{3} |
| 449 |
% + \NumberedBox{4} |
| 450 |
% + \NumberedBox{5} |
| 451 |
% + \NumberedBox{6} |
| 452 |
% + \NumberedBox{7} |
| 453 |
% + \NumberedBox{8} |
| 454 |
% + \NumberedBox{9} |
| 455 |
% + \NumberedBox{10} . |
| 456 |
% \end{autobreak} |
| 457 |
% \end{align} |
| 458 |
% \end{example} |
| 459 |
% For a technical reason, it often requires more than one run of |
| 460 |
% \LaTeX{}, and in such cases one will get informed by the following |
| 461 |
% warning: |
| 462 |
% \begin{example*} |
| 463 |
% Package autobreak Warning: Layout may have changed. |
| 464 |
% (autobreak) Rerun to get layout correct. |
| 465 |
% \end{example*} |
| 466 |
% In the next run, the layout of the equations will be corrected. |
| 467 |
% |
| 468 |
% \begin{decl} |
| 469 |
% |\MoveEqLeft|\oarg{number} |
| 470 |
% \end{decl} |
| 471 |
% This command is desinged to work like \cs{MoveEqLeft} of |
| 472 |
% the \pkg{mathtools} package^^A |
| 473 |
% \footnote{^^A |
| 474 |
% \url{https://www.ctan.org/pkg/mathtools}. |
| 475 |
% }. If it is put at the beginning of an \env{autobreak} environment, |
| 476 |
% then all subsequent lines after the first line are indented by 2 em |
| 477 |
% (the default value). |
| 478 |
% \begin{example} |
| 479 |
% \begin{align} |
| 480 |
% \begin{autobreak} |
| 481 |
% \MoveEqLeft |
| 482 |
% (n_1+n_2+n_3+n_4)^3 = |
| 483 |
% n_1^3 |
| 484 |
% + 3 n_1^2 n_2 |
| 485 |
% + 3 n_1 n_2^2 |
| 486 |
% + n_2^3 |
| 487 |
% + 3 n_1^2 n_3 |
| 488 |
% + 6 n_1 n_2 n_3 |
| 489 |
% + 3 n_2^2 n_3 |
| 490 |
% + 3 n_1 n_3^2 |
| 491 |
% + 3 n_2 n_3^2 |
| 492 |
% + n_3^3 |
| 493 |
% + 3 n_1^2 n_4 |
| 494 |
% + 6 n_1 n_2 n_4 |
| 495 |
% + 3 n_2^2 n_4 |
| 496 |
% + 6 n_1 n_3 n_4 |
| 497 |
% + 6 n_2 n_3 n_4 |
| 498 |
% + 3 n_3^2 n_4 |
| 499 |
% + 3 n_1 n_4^2 |
| 500 |
% + 3 n_2 n_4^2 |
| 501 |
% + 3 n_3 n_4^2 |
| 502 |
% + n_4^3 . |
| 503 |
% \end{autobreak} |
| 504 |
% \end{align} |
| 505 |
% \end{example} |
| 506 |
% The indent width can be changed by an optional argument of |
| 507 |
% the command. |
| 508 |
% |
| 509 |
% \begin{decl} |
| 510 |
% |\everybeforeautobreak| \marg{tokens} \\ |
| 511 |
% |\everyaftereautobreak| \marg{tokens} |
| 512 |
% \end{decl} |
| 513 |
% \SpecialUsageIndex{\everybeforeautobreak} |
| 514 |
% \SpecialUsageIndex{\everyafterautobreak} |
| 515 |
% They specify token lists inserted before and after automatically |
| 516 |
% inserted line breaks in \env{autobreak}. For example, |
| 517 |
% \begin{example**} |
| 518 |
% \begin{align} |
| 519 |
% \everyafterautobreak{\times} |
| 520 |
% \begin{autobreak} |
| 521 |
% \cos\left(\frac{\pi x}{2}\right) = |
| 522 |
% \left(1-x^2\right) |
| 523 |
% \left(1-\frac{x^2}{9}\right) |
| 524 |
% \left(1-\frac{x^2}{25}\right) |
| 525 |
% \left(1-\frac{x^2}{49}\right) |
| 526 |
% \left(1-\frac{x^2}{81}\right) |
| 527 |
% \left(1-\frac{x^2}{121}\right) |
| 528 |
% \left(1-\frac{x^2}{169}\right) |
| 529 |
% \left(1-\frac{x^2}{225}\right) |
| 530 |
% \left(1-\frac{x^2}{289}\right) |
| 531 |
% \left(1-\frac{x^2}{361}\right) |
| 532 |
% \left(1-\frac{x^2}{441}\right) |
| 533 |
% \dots |
| 534 |
% \end{autobreak} |
| 535 |
% \end{align} |
| 536 |
% \end{example**} |
| 537 |
% |
| 538 |
% \section{Caveats} |
| 539 |
% |
| 540 |
% Because \env{autobreak} tries to insert line breaks at any of new line |
| 541 |
% characters, you must not make any new lines at which the line cannot |
| 542 |
% be broken^^A |
| 543 |
% \footnote{^^A |
| 544 |
% You may put "\%" at the end of the line to avoid a new line. |
| 545 |
% }. |
| 546 |
% For example |
| 547 |
% \begin{example*} |
| 548 |
% \begin{align} |
| 549 |
% \begin{autobreak} |
| 550 |
% x = |
| 551 |
% % A problematic line break. |
| 552 |
% \frac{1} |
| 553 |
% {2} . |
| 554 |
% \end{autobreak} |
| 555 |
% \end{align} |
| 556 |
% \end{example*} |
| 557 |
% gives an error in the typesetting: |
| 558 |
% \begin{example*} |
| 559 |
% ! Missing } inserted. |
| 560 |
% <inserted text> |
| 561 |
% } |
| 562 |
% l.8 \end{align} |
| 563 |
% \end{example*} |
| 564 |
% |
| 565 |
% Putting `|\\|' or `|&|' inside \env{autobreak}, which tries to insert |
| 566 |
% these special stuffs automatically, also causes typesetting errors. |
| 567 |
% |
| 568 |
% The \env{autobreak} environment uses \cs{linewidth} as the maximum |
| 569 |
% width that expressions in its body can occupy. There is no way for |
| 570 |
% \env{autobreak} to know how much other formulae consume the space |
| 571 |
% outside it. Therefore it fails to determine the adequate maximum |
| 572 |
% width when there are expressions outside \env{autobreak} and |
| 573 |
% then \LaTeX{} gives overfull line warnings: |
| 574 |
% \begin{example*} |
| 575 |
% \begin{align} |
| 576 |
% \text{some stuff outside autobreak} |
| 577 |
% \begin{autobreak} |
| 578 |
% \text{LHS} = |
| 579 |
% ... |
| 580 |
% \end{autobreak} . % Even just a "." can be problematic. |
| 581 |
% \end{align} |
| 582 |
% % May give overfull line warnings |
| 583 |
% \end{example*} |
| 584 |
% |
| 585 |
% One may want to separate long formulae from the main document file to |
| 586 |
% other files and include them via \cs{input}\marg{file}, for example, |
| 587 |
% \begin{example*} |
| 588 |
% \begin{align} |
| 589 |
% \begin{autobreak} |
| 590 |
% \input{longeqn.inc} % It works! |
| 591 |
% \end{autobreak} |
| 592 |
% \end{align} |
| 593 |
% |
| 594 |
% \begin{align} |
| 595 |
% \begin{autobreak} |
| 596 |
% lhs = |
| 597 |
% \input{longrhs.inc} % It also works! |
| 598 |
% . |
| 599 |
% \end{autobreak} |
| 600 |
% \end{align} |
| 601 |
% \end{example*} |
| 602 |
% The current version of \env{autobreak} supports these cases: the file |
| 603 |
% content of \cs{input}\marg{file} is expanded before recognizing the |
| 604 |
% lines, with the help of the \pkg{catchfile} package^^A |
| 605 |
% \footnote{^^A |
| 606 |
% \url{https://www.ctan.org/pkg/catchfile}. |
| 607 |
% }, |
| 608 |
% when it appears at the beginning of each line. But it does not support |
| 609 |
% \cs{input}\marg{file} in the middle of the lines: |
| 610 |
% \begin{example*} |
| 611 |
% \begin{align} |
| 612 |
% \begin{autobreak} |
| 613 |
% x + \input{longexpr.inc} % Sorry, it does not work. |
| 614 |
% \end{autobreak} |
| 615 |
% \end{align} |
| 616 |
% \end{example*} |
| 617 |
% The difficulty comes from the fact that it needs to be expanded before |
| 618 |
% \env{autobreak} scans lines. By the same reason, \env{autobreak} fails |
| 619 |
% to detect new lines defined inside macros^^A |
| 620 |
% \footnote{^^A |
| 621 |
% Actually, when the definition of \cs{foo} is parsed, the new line |
| 622 |
% characters inside it are usually lost. |
| 623 |
% }: |
| 624 |
% \begin{example*} |
| 625 |
% \newcommand{\foo}{ |
| 626 |
% a |
| 627 |
% + b |
| 628 |
% + c |
| 629 |
% + d |
| 630 |
% } |
| 631 |
% \begin{align} |
| 632 |
% \begin{autobreak} |
| 633 |
% \foo + \foo + \foo + \foo % No new lines can be detected. |
| 634 |
% \end{autobreak} |
| 635 |
% \end{align} |
| 636 |
% \end{example*} |
| 637 |
% |
| 638 |
% \StopEventually{^^A |
| 639 |
% \PrintChanges |
| 640 |
% \PrintIndex |
| 641 |
% } |
| 642 |
% |
| 643 |
% \section{Implementation} |
| 644 |
% |
| 645 |
% \begin{macrocode} |
| 646 |
%<*package> |
| 647 |
\NeedsTeXFormat{LaTeX2e} |
| 648 |
\ProvidesPackage{autobreak}% |
| 649 |
[2017/02/23 v0.3 simple line breaking of long formulae] |
| 650 |
% \end{macrocode} |
| 651 |
% |
| 652 |
% \subsection{Registers and constants} |
| 653 |
% |
| 654 |
% \begin{macro}{\everybeforeautobreak} |
| 655 |
% The list of tokens that gets inserted before every line break generated |
| 656 |
% by \env{autobreak}. |
| 657 |
% \begin{macrocode} |
| 658 |
\newtoks\everybeforeautobreak |
| 659 |
% \end{macrocode} |
| 660 |
% \end{macro} |
| 661 |
% |
| 662 |
% \begin{macro}{\everyafterautobreak} |
| 663 |
% The list of tokens that gets inserted after every line break generated |
| 664 |
% by \env{autobreak}. |
| 665 |
% \begin{macrocode} |
| 666 |
\newtoks\everyafterautobreak |
| 667 |
% \end{macrocode} |
| 668 |
% \end{macro} |
| 669 |
% |
| 670 |
% \begin{macro}{\@autobreak@alltoks} |
| 671 |
% The token register to store the whole result of \env{autobreak}. |
| 672 |
% \begin{macrocode} |
| 673 |
\newtoks\@autobreak@alltoks |
| 674 |
% \end{macrocode} |
| 675 |
% \end{macro} |
| 676 |
% |
| 677 |
% \begin{macro}{\@autobreak@linetoks} |
| 678 |
% The token register for the current line. |
| 679 |
% \begin{macrocode} |
| 680 |
\newtoks\@autobreak@linetoks |
| 681 |
% \end{macrocode} |
| 682 |
% \end{macro} |
| 683 |
% |
| 684 |
% \begin{macro}{\@autobreak@lhswidth} |
| 685 |
% The width of the current left-hand side. |
| 686 |
% \begin{macrocode} |
| 687 |
\newdimen\@autobreak@lhswidth |
| 688 |
% \end{macrocode} |
| 689 |
% \end{macro} |
| 690 |
% |
| 691 |
% \begin{macro}{\@autobreak@rhswidth} |
| 692 |
% The width of the current right-hand side. |
| 693 |
% \begin{macrocode} |
| 694 |
\newdimen\@autobreak@rhswidth |
| 695 |
% \end{macrocode} |
| 696 |
% \end{macro} |
| 697 |
% |
| 698 |
% \begin{macro}{\@autobreak@maxlhswidth} |
| 699 |
% The width of the longest left-hand side occupied. Affected by the |
| 700 |
% |.aux| file generated in the previous run. |
| 701 |
% \begin{macrocode} |
| 702 |
\newdimen\@autobreak@maxlhswidth |
| 703 |
% \end{macrocode} |
| 704 |
% \end{macro} |
| 705 |
% |
| 706 |
% \begin{macro}{\@autobreak@realmaxlhswidth} |
| 707 |
% The width of the longest left-hand side occupied. Not affected by the |
| 708 |
% |.aux| file generated in the previous run. |
| 709 |
% \begin{macrocode} |
| 710 |
\newdimen\@autobreak@realmaxlhswidth |
| 711 |
% \end{macrocode} |
| 712 |
% \end{macro} |
| 713 |
% |
| 714 |
% \begin{macro}{\@autobreak@maxrhswidth} |
| 715 |
% The maximum width that the right-hand sides can occupy. |
| 716 |
% \begin{macrocode} |
| 717 |
\newdimen\@autobreak@maxrhswidth |
| 718 |
% \end{macrocode} |
| 719 |
% \end{macro} |
| 720 |
% |
| 721 |
% \begin{macro}{\@autobreak@newlinechar} |
| 722 |
% The macro representing an active |^^M|. |
| 723 |
% \begin{macrocode} |
| 724 |
\begingroup |
| 725 |
\catcode`\^^M=\active |
| 726 |
\gdef\@autobreak@newlinechar{^^M} |
| 727 |
\endgroup |
| 728 |
% \end{macrocode} |
| 729 |
% \end{macro} |
| 730 |
% |
| 731 |
% \subsection{Interaction with \texttt{.aux} files} |
| 732 |
% |
| 733 |
% When there are two or more \env{autobreak} in one \env{align}, each |
| 734 |
% \env{autobreak} has to know the maximum width of the left-hand side of |
| 735 |
% the all \env{autobreak} in the same \env{align}. Instead of violating |
| 736 |
% `causality' (e.g., how \LaTeX{} parses a file from the beginning to |
| 737 |
% the end), we use |.aux| file to store the maximum left-hand side |
| 738 |
% width, which provides the correct value in the next run. |
| 739 |
% |
| 740 |
% \begin{macro}{\if@autobreak@invalidlayout} |
| 741 |
% The switch to be turned on when an invalid layout is detected. |
| 742 |
% \begin{macrocode} |
| 743 |
\newif\if@autobreak@invalidlayout |
| 744 |
\@autobreak@invalidlayoutfalse |
| 745 |
% \end{macrocode} |
| 746 |
% \end{macro} |
| 747 |
% Show a warning if the user needs to rerun. |
| 748 |
% \begin{macrocode} |
| 749 |
\AtEndDocument{% |
| 750 |
\if@autobreak@invalidlayout |
| 751 |
\if@filesw |
| 752 |
\PackageWarningNoLine{autobreak}{Layout may have changed. |
| 753 |
\MessageBreak Rerun to get layout correct}% |
| 754 |
\else |
| 755 |
\PackageWarningNoLine{autobreak}{Layout may be wrong}% |
| 756 |
\fi |
| 757 |
\fi |
| 758 |
} |
| 759 |
% \end{macrocode} |
| 760 |
% \begin{macro}{\@autobreak@getmaxlhswidth} |
| 761 |
% To be expanded to a value saved in |.aux| in the previous run, or |
| 762 |
% |0pt| if not found. |
| 763 |
% \begin{macrocode} |
| 764 |
\def\@autobreak@getmaxlhswidth#1{% |
| 765 |
\@ifundefined{@autobreak@w@#1}{% |
| 766 |
\z@ |
| 767 |
}{% |
| 768 |
\@nameuse{@autobreak@w@#1}% |
| 769 |
}% |
| 770 |
} |
| 771 |
% \end{macrocode} |
| 772 |
% \end{macro} |
| 773 |
% |
| 774 |
% \begin{macro}{\@autobreak@setmaxlhswidth} |
| 775 |
% Called in |.aux|. |
| 776 |
% \begin{macrocode} |
| 777 |
\def\@autobreak@setmaxlhswidth#1#2{% |
| 778 |
\global\@namedef{@autobreak@w@#1}{#2}% |
| 779 |
} |
| 780 |
% \end{macrocode} |
| 781 |
% \end{macro} |
| 782 |
% |
| 783 |
% \begin{macro}{@autobreak@eqnindex} |
| 784 |
% The counter to identify each \env{align}. |
| 785 |
% \begin{macrocode} |
| 786 |
\newcounter{@autobreak@eqnindex} |
| 787 |
% \end{macrocode} |
| 788 |
% \end{macro} |
| 789 |
% |
| 790 |
% \begin{macro}{@autobreak@subeqnindex} |
| 791 |
% The counter to store the number of \env{autobreak} in an \env{align}. |
| 792 |
% \begin{macrocode} |
| 793 |
\newcounter{@autobreak@subeqnindex}[@autobreak@eqnindex]% |
| 794 |
% \end{macrocode} |
| 795 |
% \end{macro} |
| 796 |
% |
| 797 |
% \begin{macro}{\@autobreak@loadmaxlhswidth} |
| 798 |
% Loads \cs{@autobreak@maxlhswidth} for the next \env{align}. |
| 799 |
% \begin{macrocode} |
| 800 |
\def\@autobreak@loadmaxlhswidth{% |
| 801 |
\stepcounter{@autobreak@eqnindex}% |
| 802 |
\@autobreak@maxlhswidth=% |
| 803 |
\@autobreak@getmaxlhswidth{\arabic{@autobreak@eqnindex}}% |
| 804 |
\@autobreak@realmaxlhswidth=\z@ |
| 805 |
} |
| 806 |
% \end{macrocode} |
| 807 |
% \end{macro} |
| 808 |
% |
| 809 |
% \begin{macro}{\@autobreak@savemaxlhswidth} |
| 810 |
% Saves \cs{@autobreak@realmaxlhswidth} for the next run. |
| 811 |
% \begin{macrocode} |
| 812 |
\def\@autobreak@savemaxlhswidth{% |
| 813 |
\ifnum\arabic{@autobreak@subeqnindex}>0 |
| 814 |
\ifdim\@autobreak@maxlhswidth=\@autobreak@realmaxlhswidth |
| 815 |
\else |
| 816 |
% \end{macrocode} |
| 817 |
% We have used the wrong value of \cs{@autobreak@maxlhswidth} (was too |
| 818 |
% much). Need to rerun. |
| 819 |
% \begin{macrocode} |
| 820 |
\global\@autobreak@invalidlayouttrue |
| 821 |
\fi |
| 822 |
% \end{macrocode} |
| 823 |
% Note that \cs{@autobreak@maxlhswidth} becomes problematic only when |
| 824 |
% two or more \env{autobreak} appear in one \env{align}. In the case |
| 825 |
% with one \env{autobreak}, the default value |0pt| is safe for the next |
| 826 |
% run. |
| 827 |
% \begin{macrocode} |
| 828 |
\ifnum\arabic{@autobreak@subeqnindex}>1 |
| 829 |
\if@filesw |
| 830 |
% \end{macrocode} |
| 831 |
% We should provide \cs{@autobreak@setmaxlhswidth} in |.aux|. |
| 832 |
% \begin{macrocode} |
| 833 |
\@ifundefined{@autobreak@auxinited}{% |
| 834 |
\immediate\write\@mainaux{% |
| 835 |
\string\providecommand |
| 836 |
\string\@autobreak@setmaxlhswidth[2]{}% |
| 837 |
}% |
| 838 |
\gdef\@autobreak@auxinited{}% |
| 839 |
}{}% |
| 840 |
\immediate\write\@auxout{% |
| 841 |
\string\@autobreak@setmaxlhswidth% |
| 842 |
{\arabic{@autobreak@eqnindex}}% |
| 843 |
{\the\@autobreak@realmaxlhswidth}% |
| 844 |
}% |
| 845 |
\fi |
| 846 |
\fi |
| 847 |
\fi |
| 848 |
} |
| 849 |
% \end{macrocode} |
| 850 |
% \end{macro} |
| 851 |
% |
| 852 |
% \subsection{Hacking \pkg{amsmath}} |
| 853 |
% |
| 854 |
% \begin{macrocode} |
| 855 |
\RequirePackage{amsmath} |
| 856 |
% \end{macrocode} |
| 857 |
% |
| 858 |
% \begin{macro}{\if@autobreak@newlinedef} |
| 859 |
% The switch to be turned on when \cs{@autobreak@newlinedef} applies. |
| 860 |
% \begin{macrocode} |
| 861 |
\newif\if@autobreak@newlinedef |
| 862 |
\@autobreak@newlinedeffalse |
| 863 |
% \end{macrocode} |
| 864 |
% \end{macro} |
| 865 |
% |
| 866 |
% \begin{macro}{\@autobreak@newlinedef} |
| 867 |
% Installs the definition of |^^M| as a space. This is virtually |
| 868 |
% harmless in math mode. |
| 869 |
% \begin{macrocode} |
| 870 |
\begingroup |
| 871 |
\catcode`\^^M=\active |
| 872 |
\gdef\@autobreak@newlinedef{% |
| 873 |
\def^^M{ }% |
| 874 |
\@autobreak@newlinedeftrue |
| 875 |
} |
| 876 |
\endgroup |
| 877 |
% \end{macrocode} |
| 878 |
% \end{macro} |
| 879 |
% |
| 880 |
% \begin{macro}{\collect@body} |
| 881 |
% We need to override \cs{collect@body} such that it keeps |^^M|. |
| 882 |
% \begin{macrocode} |
| 883 |
\def\collect@body#1{% |
| 884 |
\@envbody={\expandafter#1\expandafter{\the\@envbody}}% |
| 885 |
\edef\process@envbody{% |
| 886 |
\the\@envbody\noexpand\end{\@currenvir}% |
| 887 |
}% |
| 888 |
\@envbody=\@emptytoks |
| 889 |
\def\begin@stack{b}% |
| 890 |
\begingroup |
| 891 |
% \end{macrocode} |
| 892 |
% Actually, the following three lines need to be inserted to the |
| 893 |
% original code. |
| 894 |
% \begin{macrocode} |
| 895 |
\if@autobreak@newlinedef |
| 896 |
\catcode`\^^M=\active |
| 897 |
\fi |
| 898 |
\expandafter\let\csname\@currenvir\endcsname=\collect@@body |
| 899 |
\edef\process@envbody{% |
| 900 |
\expandafter\noexpand\csname\@currenvir\endcsname |
| 901 |
}% |
| 902 |
\process@envbody |
| 903 |
} |
| 904 |
% \end{macrocode} |
| 905 |
% \end{macro} |
| 906 |
% |
| 907 |
% \begin{environment}{align} |
| 908 |
% Hack \env{align} of \pkg{amsmath}. |
| 909 |
% \begin{macrocode} |
| 910 |
\let\@autobreak@oldstart@align=\start@align |
| 911 |
\def\start@align{% |
| 912 |
\@autobreak@loadmaxlhswidth |
| 913 |
\@autobreak@newlinedef |
| 914 |
\@autobreak@oldstart@align |
| 915 |
} |
| 916 |
% \end{macrocode} |
| 917 |
% \begin{macrocode} |
| 918 |
\let\@autobreak@oldendalign=\endalign |
| 919 |
\def\endalign{% |
| 920 |
\@autobreak@savemaxlhswidth |
| 921 |
\@autobreak@oldendalign |
| 922 |
} |
| 923 |
% \end{macrocode} |
| 924 |
% \end{environment} |
| 925 |
% |
| 926 |
% \subsection{\env{autobreak} environment} |
| 927 |
% |
| 928 |
% \begin{environment}{autobreak} |
| 929 |
% Checks if we are in \env{align} (and \cs{@autobreak@newlinedef} is |
| 930 |
% applied), increments the counter and collects its body via |
| 931 |
% \cs{collect@body}. |
| 932 |
% \begin{macrocode} |
| 933 |
\newenvironment{autobreak}{% |
| 934 |
\if@autobreak@newlinedef |
| 935 |
\else |
| 936 |
\PackageError{autobreak}{% |
| 937 |
autobreak is not allowed here% |
| 938 |
}{% |
| 939 |
Use autobreak inside align. |
| 940 |
}% |
| 941 |
\fi |
| 942 |
\stepcounter{@autobreak@subeqnindex}% |
| 943 |
\collect@body\@autobreak |
| 944 |
}{} |
| 945 |
% \end{macrocode} |
| 946 |
% \end{environment} |
| 947 |
% |
| 948 |
% \begin{macro}{\@autobreak} |
| 949 |
% Called from \cs{collect@body}. The parameter |#1| is the whole body. |
| 950 |
% It takes also |#2| and |#3|, which are always \cs{end} and |
| 951 |
% |autobreak|, to remove them from the successive tokens. |
| 952 |
% \begin{macrocode} |
| 953 |
\def\@autobreak#1#2#3{% |
| 954 |
% \end{macrocode} |
| 955 |
% First, close the group of \env{autobreak}. |
| 956 |
% \begin{macrocode} |
| 957 |
\end{autobreak}% |
| 958 |
% \end{macrocode} |
| 959 |
% Then parse the given body of the environment and construct lines to be |
| 960 |
% passed to \env{align}. |
| 961 |
% \begin{macrocode} |
| 962 |
\@autobreak@init |
| 963 |
\def\@tempa{\expandafter\@autobreak@scanline |
| 964 |
\@autobreak@newlinechar#1}% |
| 965 |
\expandafter\@tempa\@autobreak@newlinechar\@autobreak@end |
| 966 |
} |
| 967 |
% \end{macrocode} |
| 968 |
% \end{macro} |
| 969 |
% |
| 970 |
% \begin{macro}{\@autobreak@init} |
| 971 |
% Initialization. |
| 972 |
% \begin{macrocode} |
| 973 |
\def\@autobreak@init{% |
| 974 |
\@autobreak@alltoks={}% |
| 975 |
\@autobreak@linetoks={}% |
| 976 |
\@autobreak@lhswidth=\z@ |
| 977 |
\let\MoveEqLeft=\@autobreak@MoveEqLeft |
| 978 |
} |
| 979 |
% \end{macrocode} |
| 980 |
% \end{macro} |
| 981 |
% |
| 982 |
% \begin{macro}{\@autobreak@end} |
| 983 |
% Finalization. It generates the whole lines in one go. |
| 984 |
% \begin{macrocode} |
| 985 |
\def\@autobreak@end{% |
| 986 |
\expandafter\@autobreak@addtoks\expandafter\@autobreak@alltoks |
| 987 |
\expandafter{\the\@autobreak@linetoks}% |
| 988 |
\the\@autobreak@alltoks |
| 989 |
} |
| 990 |
% \end{macrocode} |
| 991 |
% \end{macro} |
| 992 |
% |
| 993 |
% \begin{macro}{\@autobreak@scanline} |
| 994 |
% Takes a line from the input stream. Here a line ends with |^^M|. |
| 995 |
% \begin{macrocode} |
| 996 |
\begingroup |
| 997 |
\catcode`\^^M=\active |
| 998 |
\gdef\@autobreak@scanline#1^^M{\@autobreak@scanline@{#1}} |
| 999 |
\endgroup |
| 1000 |
% \end{macrocode} |
| 1001 |
% If the next token is a punctuation, then we merge it into the current |
| 1002 |
% line. (Otherwise it can make a line only with a period, for example). |
| 1003 |
% \begin{macrocode} |
| 1004 |
\def\@autobreak@scanline@#1{% |
| 1005 |
\@autobreak@ifnextpunct{% |
| 1006 |
\@autobreak@scanline@gobble{#1}% |
| 1007 |
}{% |
| 1008 |
\@autobreak@scanline@@{#1}% |
| 1009 |
}% |
| 1010 |
} |
| 1011 |
% \end{macrocode} |
| 1012 |
% A helper macro of \cs{@ifnextpunct}\marg{if-yes}\marg{if-no}. |
| 1013 |
% \begin{macrocode} |
| 1014 |
\def\@autobreak@ifnextpunct#1#2{% |
| 1015 |
\@ifnextchar.{% |
| 1016 |
#1% |
| 1017 |
}{% |
| 1018 |
\@ifnextchar,{% |
| 1019 |
#1% |
| 1020 |
}{% |
| 1021 |
\@ifnextchar;{% |
| 1022 |
#1% |
| 1023 |
}{% |
| 1024 |
\@ifnextchar:{% |
| 1025 |
#1% |
| 1026 |
}{% |
| 1027 |
#2% |
| 1028 |
}% |
| 1029 |
}% |
| 1030 |
}% |
| 1031 |
}% |
| 1032 |
} |
| 1033 |
% \end{macrocode} |
| 1034 |
% Merge punctuations as possible (usually there is only one period in a |
| 1035 |
% line, though). |
| 1036 |
% \begin{macrocode} |
| 1037 |
\def\@autobreak@scanline@gobble#1#2{% |
| 1038 |
\@autobreak@ifnextpunct{% |
| 1039 |
\@autobreak@scanline@gobble{#1#2}% |
| 1040 |
}{% |
| 1041 |
\@autobreak@scanline@@{#1#2}% |
| 1042 |
}% |
| 1043 |
} |
| 1044 |
% \end{macrocode} |
| 1045 |
% Pass the current line to \cs{@autobreak@processline}. Then, repeat |
| 1046 |
% scanning lines until \cs{@autobreak@end} appears as the next token. |
| 1047 |
% \begin{macrocode} |
| 1048 |
\def\@autobreak@scanline@@#1{% |
| 1049 |
\@autobreak@processline{#1}% |
| 1050 |
\@ifnextchar\@autobreak@end{}{% |
| 1051 |
\@autobreak@scanline@@@ |
| 1052 |
}% |
| 1053 |
} |
| 1054 |
% \end{macrocode} |
| 1055 |
% Catch \cs{MoveEqLeft}. |
| 1056 |
% \changes{v0.3}{2017/02/23}{Add \cs{MoveEqLeft} command} |
| 1057 |
% \begin{macrocode} |
| 1058 |
\def\@autobreak@scanline@@@{% |
| 1059 |
\@ifnextchar\MoveEqLeft{% |
| 1060 |
\@autobreak@scanline@MoveEqLeft |
| 1061 |
}{% |
| 1062 |
\@autobreak@scanline@@@@ |
| 1063 |
}% |
| 1064 |
} |
| 1065 |
% \end{macrocode} |
| 1066 |
% The argument |#1| is \cs{MoveEqLeft}. This command accepts an |
| 1067 |
% optional number. |
| 1068 |
% \begin{macrocode} |
| 1069 |
\def\@autobreak@scanline@MoveEqLeft#1{% |
| 1070 |
\@ifnextchar[{% |
| 1071 |
\@autobreak@scanline@MoveEqLeft@ |
| 1072 |
}{% |
| 1073 |
\@autobreak@scanline@MoveEqLeft@[2]% |
| 1074 |
}% |
| 1075 |
} |
| 1076 |
% \end{macrocode} |
| 1077 |
% \begin{macrocode} |
| 1078 |
\def\@autobreak@scanline@MoveEqLeft@[#1]{% |
| 1079 |
\ifdim#1\p@>\z@ |
| 1080 |
\def\@tempa{\@autobreak@scanline@MvEqL@pos}% |
| 1081 |
\else\ifdim#1\p@=\z@ |
| 1082 |
\def\@tempa{\@autobreak@scanline@MvEqL@zero}% |
| 1083 |
\else |
| 1084 |
\def\@tempa{\@autobreak@scanline@MvEqL@neg}% |
| 1085 |
\fi\fi |
| 1086 |
\@tempa{#1}% |
| 1087 |
} |
| 1088 |
% \end{macrocode} |
| 1089 |
% \begin{macrocode} |
| 1090 |
\def\@autobreak@scanline@MvEqL@pos#1{% |
| 1091 |
\def\@tempa{\expandafter\@autobreak@scanline\kern#1em}% |
| 1092 |
\expandafter\@tempa\@autobreak@newlinechar\kern-#1em% |
| 1093 |
} |
| 1094 |
% \end{macrocode} |
| 1095 |
% In the case with |#1| = 0, a special treatment is required because \\ |
| 1096 |
% \cs{@autobreak@processline} ignores a zero width. Insert a very tiny |
| 1097 |
% space. |
| 1098 |
% \begin{macrocode} |
| 1099 |
\def\@autobreak@scanline@MvEqL@zero#1{% |
| 1100 |
\def\@tempa{\expandafter\@autobreak@scanline\kern1sp}% |
| 1101 |
\expandafter\@tempa\@autobreak@newlinechar\kern-1sp% |
| 1102 |
} |
| 1103 |
% \end{macrocode} |
| 1104 |
% In the case with |#1| < 0, put a very tiny space, and then put the |
| 1105 |
% space with a positive width such that the first line is indented to |
| 1106 |
% the right. |
| 1107 |
% \begin{macrocode} |
| 1108 |
\def\@autobreak@scanline@MvEqL@neg#1{% |
| 1109 |
\def\@tempa{\expandafter\@autobreak@scanline\kern1sp}% |
| 1110 |
\expandafter\@tempa\@autobreak@newlinechar\kern-1sp\kern-#1em% |
| 1111 |
} |
| 1112 |
% \end{macrocode} |
| 1113 |
% One may expect \cs{input}\marg{file} in \env{autobreak} is expanded by |
| 1114 |
% the file content and \env{autobreak} treats new lines in it correctly. |
| 1115 |
% But it needs more work. Because handling of \cs{input} in the middle |
| 1116 |
% of the lines is rather involved, for now we support only \cs{input} at |
| 1117 |
% the beginning of each line (which is what sane people usually do). |
| 1118 |
% This can be done via the \pkg{catchfile} package. |
| 1119 |
% \begin{macrocode} |
| 1120 |
\IfFileExists{catchfile.sty}{ |
| 1121 |
\RequirePackage{catchfile} |
| 1122 |
\def\@autobreak@scanline@@@@{% |
| 1123 |
\@ifnextchar\input{% |
| 1124 |
\@autobreak@scanline@input |
| 1125 |
}{% |
| 1126 |
\@autobreak@scanline |
| 1127 |
}% |
| 1128 |
}% |
| 1129 |
}{ |
| 1130 |
\def\@autobreak@scanline@@@@{% |
| 1131 |
\@ifnextchar\input{% |
| 1132 |
\PackageWarning{autobreak}{% |
| 1133 |
Cannot handle new lines in a file via \protect\input, |
| 1134 |
\MessageBreak which requires the catchfile package |
| 1135 |
}% |
| 1136 |
}% |
| 1137 |
\@autobreak@scanline |
| 1138 |
}% |
| 1139 |
} |
| 1140 |
% \end{macrocode} |
| 1141 |
% The argument |#1| is \cs{input} and |#2| is the file name. |
| 1142 |
% \begin{macrocode} |
| 1143 |
\def\@autobreak@scanline@input#1#2{% |
| 1144 |
\CatchFileDef\@tempa{#2}{\catcode`\^^M=\active}% |
| 1145 |
\expandafter\@autobreak@scanline\@tempa |
| 1146 |
} |
| 1147 |
% \end{macrocode} |
| 1148 |
% \end{macro} |
| 1149 |
% |
| 1150 |
% \begin{macro}{\@autobreak@MoveEqLeft} |
| 1151 |
% This definition is expanded only when \cs{@autobreak@scanline} cannot |
| 1152 |
% detect \cs{MoveEqLeft} in an \env{autobreak} environment, in other |
| 1153 |
% words, when it appears in the middle of a line. |
| 1154 |
% \begin{macrocode} |
| 1155 |
\def\@autobreak@MoveEqLeft{% |
| 1156 |
\PackageError{autobreak}{% |
| 1157 |
\protect\MoveEqLeft\space is not allowed here% |
| 1158 |
}{% |
| 1159 |
\protect\MoveEqLeft\space must be put at the beginning of |
| 1160 |
an autobreak environment. |
| 1161 |
}% |
| 1162 |
} |
| 1163 |
% \end{macrocode} |
| 1164 |
% \end{macro} |
| 1165 |
% |
| 1166 |
% \begin{macro}{\@autobreak@processline} |
| 1167 |
% \changes{v0.2}{2016/07/03}{Fix space calculation around alignment |
| 1168 |
% tabs} |
| 1169 |
% Each line from \cs{autobreak@scanline} should be regarded as a `block' |
| 1170 |
% in the equation. The first block (typically the left-hand side $+$ |
| 1171 |
% `=') determines the indentation for the successive lines. From the |
| 1172 |
% second block, try to append the block to the end of the line and |
| 1173 |
% insert a line break if it does not fit in a line. Note that we measure |
| 1174 |
% the widths of the blocks with putting |{}| around alignment tabs. |
| 1175 |
% \begin{macrocode} |
| 1176 |
\def\@autobreak@processline#1{% |
| 1177 |
\ifdim\@autobreak@lhswidth=\z@ |
| 1178 |
% \end{macrocode} |
| 1179 |
% For the first block. The rest of the width for the right-hand sides |
| 1180 |
% is determined from \cs{linewidth} and \cs{@autobreak@maxlhswidth}. |
| 1181 |
% \begin{macrocode} |
| 1182 |
\@autobreak@settowidth\@autobreak@lhswidth{#1{}}% |
| 1183 |
\ifdim\@autobreak@lhswidth>\z@ |
| 1184 |
\ifdim\@autobreak@lhswidth>\@autobreak@maxlhswidth |
| 1185 |
\ifdim\@autobreak@maxlhswidth>\z@ |
| 1186 |
% \end{macrocode} |
| 1187 |
% The previous one used the wrong value of \cs{@autobreak@maxlhswidth} |
| 1188 |
% (was too short). Need to rerun. |
| 1189 |
% \begin{macrocode} |
| 1190 |
\global\@autobreak@invalidlayouttrue |
| 1191 |
\fi |
| 1192 |
\global\@autobreak@maxlhswidth=\@autobreak@lhswidth |
| 1193 |
\fi |
| 1194 |
\ifdim\@autobreak@lhswidth>\@autobreak@realmaxlhswidth |
| 1195 |
\global\@autobreak@realmaxlhswidth=\@autobreak@lhswidth |
| 1196 |
\fi |
| 1197 |
\@autobreak@maxrhswidth=\linewidth |
| 1198 |
\advance\@autobreak@maxrhswidth by -\@autobreak@maxlhswidth |
| 1199 |
\@autobreak@alltoks={#1{}&}% |
| 1200 |
\fi |
| 1201 |
\else |
| 1202 |
% \end{macrocode} |
| 1203 |
% For the rest of the blocks. |
| 1204 |
% \begin{macrocode} |
| 1205 |
\@autobreak@settowidth\@autobreak@rhswidth |
| 1206 |
{{}\the\@autobreak@linetoks#1\the\everybeforeautobreak}% |
| 1207 |
\ifdim\@autobreak@rhswidth>\@autobreak@maxrhswidth |
| 1208 |
% \end{macrocode} |
| 1209 |
% Adding the next block gives an overfull line. Need a line break. |
| 1210 |
% \begin{macrocode} |
| 1211 |
\edef\@tempa{\the\@autobreak@linetoks\the\everybeforeautobreak}% |
| 1212 |
\expandafter\@autobreak@addtoks\expandafter\@autobreak@alltoks |
| 1213 |
\expandafter{\@tempa\notag\\&}% |
| 1214 |
\@autobreak@linetoks=\everyafterautobreak |
| 1215 |
\fi |
| 1216 |
\@autobreak@addtoks\@autobreak@linetoks{#1}% |
| 1217 |
\fi |
| 1218 |
} |
| 1219 |
% \end{macrocode} |
| 1220 |
% \end{macro} |
| 1221 |
% |
| 1222 |
% \begin{macro}{\@autobreak@addtoks} |
| 1223 |
% Appends |#2| to the token register |#1|. |
| 1224 |
% \begin{macrocode} |
| 1225 |
\def\@autobreak@addtoks#1#2{% |
| 1226 |
#1=\expandafter{\the#1#2}% |
| 1227 |
} |
| 1228 |
% \end{macrocode} |
| 1229 |
% \end{macro} |
| 1230 |
% |
| 1231 |
% \begin{macro}{\@autobreak@settowidth} |
| 1232 |
% Same as \cs{settowidth} but in math mode. We assume \cs{displaystyle}. |
| 1233 |
% (Anyway \env{align} issues \cs{displaystyle} at the beginning of every |
| 1234 |
% cell.) |
| 1235 |
% \begin{macrocode} |
| 1236 |
\def\@autobreak@settowidth#1#2{% |
| 1237 |
\settowidth#1{$\displaystyle#2$}% |
| 1238 |
} |
| 1239 |
% \end{macrocode} |
| 1240 |
% \end{macro} |
| 1241 |
% |
| 1242 |
% \begin{macrocode} |
| 1243 |
%</package> |
| 1244 |
% \end{macrocode} |
| 1245 |
% \Finale |
| 1246 |
\endinput |