| 1 |
% \iffalse meta-comment |
| 2 |
% |
| 3 |
% Copyright (C) 2000-2025 by Scott Pakin <[email protected]> |
| 4 |
% ----------------------------------------------------------- |
| 5 |
% |
| 6 |
% This file may be distributed and/or modified under the |
| 7 |
% conditions of the LaTeX Project Public License, either version 1.3a |
| 8 |
% of this license or (at your option) any later version. |
| 9 |
% The latest version of this license is in: |
| 10 |
% |
| 11 |
% http://www.latex-project.org/lppl.txt |
| 12 |
% |
| 13 |
% and version 1.3c or later is part of all distributions of LaTeX |
| 14 |
% version 2006/05/20 or later. |
| 15 |
% |
| 16 |
% \fi |
| 17 |
% |
| 18 |
% \iffalse |
| 19 |
%<*driver> |
| 20 |
\ProvidesFile{bytefield.dtx} |
| 21 |
%</driver> |
| 22 |
%<package>\NeedsTeXFormat{LaTeX2e}[1999/12/01] |
| 23 |
%<package>\ProvidesPackage{bytefield} |
| 24 |
%<*package> |
| 25 |
[2025/01/25 v2.9 Network protocol diagrams] |
| 26 |
%</package> |
| 27 |
% |
| 28 |
%<*driver> |
| 29 |
\documentclass{ltxdoc} |
| 30 |
\usepackage{textcomp} |
| 31 |
\usepackage{bytefield} |
| 32 |
\usepackage{color} |
| 33 |
\usepackage{rotating} |
| 34 |
\usepackage{multirow} |
| 35 |
\usepackage{calc} |
| 36 |
\usepackage{array} |
| 37 |
\usepackage{wasysym} |
| 38 |
\usepackage{needspace} |
| 39 |
\usepackage{hypdoc} |
| 40 |
\usepackage{hyperref} |
| 41 |
\usepackage[figure]{hypcap} |
| 42 |
\usepackage{hyperxmp} |
| 43 |
|
| 44 |
\hypersetup{% |
| 45 |
pdftitle={The bytefield package}, |
| 46 |
pdfauthor={Scott Pakin}, |
| 47 |
pdfsubject={Protocol diagrams for LaTeX}, |
| 48 |
pdfkeywords={bits, bytes, bit fields, communication, network protocol diagrams, LaTeX2e, memory maps}, |
| 49 |
pdfcopyright={Copyright (C) 2000-2025, Scott Pakin}, |
| 50 |
pdflicenseurl={http://www.latex-project.org/lppl/}, |
| 51 |
pdfcaptionwriter={Scott Pakin}, |
| 52 |
pdfcontactemail={[email protected]}, |
| 53 |
pdfcontacturl={http://www.pakin.org/\xmptilde scott/}, |
| 54 |
pdflang={en-US}, |
| 55 |
baseurl={http://mirrors.ctan.org/macros/latex/contrib/bytefield/bytefield.pdf}, |
| 56 |
pdfstartview=Fit, |
| 57 |
colorlinks=false, |
| 58 |
bookmarksopen=true} |
| 59 |
\EnableCrossrefs |
| 60 |
\CodelineIndex |
| 61 |
\RecordChanges |
| 62 |
\setcounter{IndexColumns}{2} |
| 63 |
\begin{document} |
| 64 |
\DocInput{bytefield.dtx} |
| 65 |
\PrintChanges |
| 66 |
\PrintIndex |
| 67 |
\end{document} |
| 68 |
%</driver> |
| 69 |
% \fi |
| 70 |
% |
| 71 |
% \CheckSum{1205} |
| 72 |
% |
| 73 |
% \CharacterTable |
| 74 |
% {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 |
| 75 |
% 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 |
| 76 |
% Digits \0\1\2\3\4\5\6\7\8\9 |
| 77 |
% Exclamation \! Double quote \" Hash (number) \# |
| 78 |
% Dollar \$ Percent \% Ampersand \& |
| 79 |
% Acute accent \' Left paren \( Right paren \) |
| 80 |
% Asterisk \* Plus \+ Comma \, |
| 81 |
% Minus \- Point \. Solidus \/ |
| 82 |
% Colon \: Semicolon \; Less than \< |
| 83 |
% Equals \= Greater than \> Question mark \? |
| 84 |
% Commercial at \@ Left bracket \[ Backslash \\ |
| 85 |
% Right bracket \] Circumflex \^ Underscore \_ |
| 86 |
% Grave accent \` Left brace \{ Vertical bar \| |
| 87 |
% Right brace \} Tilde \~} |
| 88 |
% |
| 89 |
% |
| 90 |
% \changes{v1.0}{2000/07/02}{Initial version} |
| 91 |
% \changes{v1.1}{2002/09/14}{Restructured the \texttt{.dtx} file} |
| 92 |
% \changes{v1.4}{2011/01/09}{Made assignments to \cs{counting@words} |
| 93 |
% global to prevent vertical-spacing problems with back-to-back word |
| 94 |
% groups (bug fix due to Steven R. King)} |
| 95 |
% \changes{v1.4}{2011/01/16}{Split \cs{curlyspace}, \cs{labelspace}, and |
| 96 |
% \cs{curlyshrinkage} into \texttt{left} and \texttt{right} versions} |
| 97 |
% \changes{v2.0}{2011/01/18}{Made a number of non-backwards-compatible |
| 98 |
% changes, including replacing \cs{wordgroupr} and \cs{endwordgroupr} |
| 99 |
% with a \texttt{rightwordgroup} environment and \cs{wordgroupl} and |
| 100 |
% \cs{endwordgroupl} with a \texttt{leftwordgroup} environment and |
| 101 |
% also replacing a slew of user-visible lengths and macros with a |
| 102 |
% single \cs{bytefieldsetup} macro} |
| 103 |
% \changes{v2.5}{2020/10/20}{Redefine the \protect\cs{bitbox}, |
| 104 |
% \protect\cs{wordbox}, and \protect\cs{bitboxes} commands additionally |
| 105 |
% to accept key/value options} |
| 106 |
% \changes{v2.5}{2020/10/20}{Accept a new \protect\optname{bgcolor} |
| 107 |
% option to set a bit's background color} |
| 108 |
% \changes{v2.5}{2020/10/22}{Accept a new \protect\optname{perword} |
| 109 |
% option to execute a macro for each word in a word box. This addresses |
| 110 |
% a feature request by Victor Toni} |
| 111 |
% \changes{v2.6}{2020/10/29}{Accept new \protect\optname{curlystyle}, |
| 112 |
% \protect\optname{leftcurlystyle}, and |
| 113 |
% \protect\optname{rightcurlystyle} options to control the styling |
| 114 |
% (e.g.,~color) of curly braces. This addresses a feature request by |
| 115 |
% Victor Toni} |
| 116 |
% \changes{v2.8}{2023/09/24}{Corrected the documentation. Thanks to {\"O}mer |
| 117 |
% Faruk Birg{\"u}l for pointing out an error in one of the examples} |
| 118 |
% |
| 119 |
% \GetFileInfo{bytefield.dtx} |
| 120 |
% |
| 121 |
% \DoNotIndex{\&,\@ifstar,\@ifundefined,\@tempcnta,\@tempdima} |
| 122 |
% \DoNotIndex{\DeclareRobustCommand,\MessageBreak,\\,\addtocounter,\advance} |
| 123 |
% \DoNotIndex{\begin,\begingroup,\bgroup,\catcode,\ch@ck,\count,\cr,\csname} |
| 124 |
% \DoNotIndex{\def,\dimen,\dimendef,\edef,\egroup,\else,\end,\endcsname} |
| 125 |
% \DoNotIndex{\endgroup,\expandafter,\fi,\gdef,\global,\hspace,\if,\ifnum} |
| 126 |
% \DoNotIndex{\ifx,\ignorespaces,\insc@unt,\left,\let,\loop,\newbox} |
| 127 |
% \DoNotIndex{\newcommand,\newcounter,\newenvironment,\newif,\newlength} |
| 128 |
% \DoNotIndex{\newsavebox,\next,\noexpand,\par,\protect,\raisebox,\ratio} |
| 129 |
% \DoNotIndex{\real,\relax,\renewcommand,\renewenvironment,\repeat,\right} |
| 130 |
% \DoNotIndex{\sbox,\setcounter,\setlength,\settowidth,\space,\string} |
| 131 |
% \DoNotIndex{\strip@pt,\the,\unskip,\usebox,\usepackage,\value,\vbox} |
| 132 |
% \DoNotIndex{\vspace,\wlog,\xdef,\{,\},\\} |
| 133 |
% |
| 134 |
% |
| 135 |
% \title{The \textsf{bytefield} package\thanks{This document |
| 136 |
% corresponds to \textsf{bytefield}~\fileversion, dated \filedate.}} |
| 137 |
% \author{Scott Pakin \\ \textit{[email protected]}} |
| 138 |
% |
| 139 |
% \sloppy |
| 140 |
% \maketitle |
| 141 |
% |
| 142 |
% ^^A Define an environment for command/environment delcarations. |
| 143 |
% \newsavebox{\declbox} |
| 144 |
% \newenvironment{decl}{^^A |
| 145 |
% \begin{lrbox}{\declbox}\begin{tabular}{l}}{^^A |
| 146 |
% \end{tabular}\end{lrbox}^^A |
| 147 |
% \vspace{3ex}\noindent\hspace*{-3em}\fbox{\usebox{\declbox}}\vspace*{2ex}} |
| 148 |
% |
| 149 |
% ^^A Define an ellipsis that permits a subsequent line break. |
| 150 |
% \newcommand{\ellipsis}{$\ldots$\linebreak[0]} |
| 151 |
% |
| 152 |
% ^^A Define an environment for showing sample bytefield diagrams. |
| 153 |
% \newenvironment{bffigure}{% |
| 154 |
% \bigskip |
| 155 |
% \noindent\hspace*{3em}% |
| 156 |
% }{% |
| 157 |
% \bigskip |
| 158 |
% } |
| 159 |
% |
| 160 |
% ^^A Define something like \cs to use within \StopEventually. |
| 161 |
% \DeclareRobustCommand{\cseq}[1]{\texttt{\string#1}} |
| 162 |
% |
| 163 |
% ^^A Index a package. This code was adapted from \SpecialEnvIndex. |
| 164 |
% \makeatletter |
| 165 |
% \def\SpecialPkgIndex#1{^^A |
| 166 |
% \@bsphack |
| 167 |
% \index{#1\actualchar{\protect\sffamily#1} (package)\encapchar usage}^^A |
| 168 |
% \index{packages:\levelchar#1\actualchar{\protect\sffamily#1}^^A |
| 169 |
% \encapchar usage}^^A |
| 170 |
% \@esphack |
| 171 |
% } |
| 172 |
% |
| 173 |
% ^^A The following was adapted from hypdoc's redefinition of \SpecialEnvIndex. |
| 174 |
% \expandafter\ifx\csname c@HD@hypercount\endcsname\relax |
| 175 |
% \else |
| 176 |
% \let\HDorg@SpecialPkgIndex\SpecialPkgIndex |
| 177 |
% \renewcommand*\SpecialPkgIndex[1]{^^A |
| 178 |
% \@bsphack |
| 179 |
% \begingroup |
| 180 |
% \HD@target |
| 181 |
% \let\HDorg@encapchar\encapchar |
| 182 |
% \edef\encapchar usage{^^A |
| 183 |
% \HDorg@encapchar hdpindex{usage}^^A |
| 184 |
% }^^A |
| 185 |
% \HDorg@SpecialPkgIndex{#1}^^A |
| 186 |
% \endgroup |
| 187 |
% \@esphack |
| 188 |
% } |
| 189 |
% \fi |
| 190 |
% \makeatother |
| 191 |
% |
| 192 |
% ^^A Define a logical style for package names. |
| 193 |
% \DeclareRobustCommand{\pkgname}[1]{^^A |
| 194 |
% \textsf{#1}^^A |
| 195 |
% \SpecialPkgIndex{#1}^^A |
| 196 |
% } |
| 197 |
% |
| 198 |
% |
| 199 |
% ^^A Index an option. This code was adapted from \SpecialEnvIndex. |
| 200 |
% \makeatletter |
| 201 |
% \def\SpecialOptionIndex#1{^^A |
| 202 |
% \@bsphack |
| 203 |
% \index{#1\actualchar{\protect\sffamily#1} (option)\encapchar usage}^^A |
| 204 |
% \index{options:\levelchar#1\actualchar{\protect\sffamily#1}^^A |
| 205 |
% \encapchar usage}^^A |
| 206 |
% \@esphack |
| 207 |
% } |
| 208 |
% |
| 209 |
% ^^A The following was adapted from hypdoc's redefinition of \SpecialEnvIndex. |
| 210 |
% \expandafter\ifx\csname c@HD@hypercount\endcsname\relax |
| 211 |
% \else |
| 212 |
% \let\HDorg@SpecialOptionIndex\SpecialOptionIndex |
| 213 |
% \renewcommand*\SpecialOptionIndex[1]{^^A |
| 214 |
% \@bsphack |
| 215 |
% \begingroup |
| 216 |
% \HD@target |
| 217 |
% \let\HDorg@encapchar\encapchar |
| 218 |
% \edef\encapchar usage{^^A |
| 219 |
% \HDorg@encapchar hdpindex{usage}^^A |
| 220 |
% }^^A |
| 221 |
% \HDorg@SpecialOptionIndex{#1}^^A |
| 222 |
% \endgroup |
| 223 |
% \@esphack |
| 224 |
% } |
| 225 |
% \fi |
| 226 |
% \makeatother |
| 227 |
% |
| 228 |
% ^^A Define a logical style for package-option names. |
| 229 |
% \DeclareRobustCommand{\optname}[1]{^^A |
| 230 |
% \texttt{#1}^^A |
| 231 |
% \SpecialOptionIndex{#1}^^A |
| 232 |
% } |
| 233 |
% |
| 234 |
% ^^A We use this box to help center protocol diagrams. |
| 235 |
% \newsavebox{\protocoldiagram} |
| 236 |
% |
| 237 |
% |
| 238 |
% \begin{abstract} |
| 239 |
% The \pkgname{bytefield} package helps the user create illustrations for |
| 240 |
% network protocol specifications and anything else that utilizes fields of |
| 241 |
% data. These illustrations show how the bits and bytes are laid out in a |
| 242 |
% packet or in memory. |
| 243 |
% \end{abstract} |
| 244 |
% |
| 245 |
% |
| 246 |
% \vspace{\baselineskip} |
| 247 |
% \begin{center} |
| 248 |
% \fbox{% |
| 249 |
% \begin{minipage}{0.65\linewidth} |
| 250 |
% \textsc{Warning}: \pkgname{bytefield} version~2.\textit{x} breaks |
| 251 |
% compatibility with older versions of the package. See |
| 252 |
% Section~\ref{sec:upgrading} for help porting documents to the new |
| 253 |
% interface. |
| 254 |
% \end{minipage}% |
| 255 |
% } |
| 256 |
% \end{center} |
| 257 |
% \vspace{\baselineskip} |
| 258 |
% |
| 259 |
% |
| 260 |
% \section{Introduction} |
| 261 |
% |
| 262 |
% Network protocols are usually specified in terms of a sequence of bits |
| 263 |
% and bytes arranged in a field. This is portrayed graphically as a |
| 264 |
% grid of boxes. Each row in the grid represents one word (frequently, |
| 265 |
% 8, 16, or 32~bits), and each column represents a bit within a word. |
| 266 |
% The \pkgname{bytefield} package makes it easy to typeset these sorts |
| 267 |
% of figures. \pkgname{bytefield} facilitates drawing protocol diagrams |
| 268 |
% that contain |
| 269 |
% |
| 270 |
% \begin{itemize} |
| 271 |
% \item words of any arbitrary number of bits, |
| 272 |
% \item column headers showing bit positions, |
| 273 |
% \item multiword fields---even non-word-aligned and even if the total |
| 274 |
% number of bits is not a multiple of the word length, |
| 275 |
% \item word labels on either the left or right of the figure, and |
| 276 |
% \item ``skipped words'' within fields. |
| 277 |
% \end{itemize} |
| 278 |
% |
| 279 |
% Because \pkgname{bytefield} draws its figures using only the \LaTeX{} |
| 280 |
% \texttt{picture} environment, these figures are not specific to any |
| 281 |
% particular backend, do not require PostScript or PDF support, and do |
| 282 |
% not need support from external programs. Furthermore, unlike an |
| 283 |
% imported graphic, \pkgname{bytefield} pictures can include arbitrary |
| 284 |
% \LaTeX{} constructs, such as mathematical equations, |\ref|s and |
| 285 |
% |\cite|s to the surrounding document, and macro calls. |
| 286 |
% |
| 287 |
% |
| 288 |
% \section{Usage} |
| 289 |
% \label{sec:usage} |
| 290 |
% |
| 291 |
% \subsection{A first example} |
| 292 |
% \label{sec:first-example} |
| 293 |
% |
| 294 |
% The Internet Engineering Task Force's Request for Comments (RFC) |
| 295 |
% number~3016 includes the following ASCII-graphics illustration of |
| 296 |
% the RTP packetization of an \mbox{MPEG-4} Visual bitstream: |
| 297 |
% |
| 298 |
% \begin{verbatim} |
| 299 |
% 0 1 2 3 |
| 300 |
% 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 |
| 301 |
% +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |
| 302 |
% |V=2|P|X| CC |M| PT | sequence number | RTP |
| 303 |
% +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |
| 304 |
% | timestamp | Header |
| 305 |
% +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |
| 306 |
% | synchronization source (SSRC) identifier | |
| 307 |
% +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ |
| 308 |
% | contributing source (CSRC) identifiers | |
| 309 |
% | .... | |
| 310 |
% +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ |
| 311 |
% | | RTP |
| 312 |
% | MPEG-4 Visual stream (byte aligned) | Pay- |
| 313 |
% | | load |
| 314 |
% | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |
| 315 |
% | :...OPTIONAL RTP padding | |
| 316 |
% +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |
| 317 |
% \end{verbatim} |
| 318 |
% |
| 319 |
% \noindent |
| 320 |
% The following \LaTeX\ code shows how straightforward it is to typeset |
| 321 |
% that illustration using the \pkgname{bytefield} package: |
| 322 |
% |
| 323 |
% \begin{verbatim} |
| 324 |
% \begin{bytefield}[bitwidth=1.1em]{32} |
| 325 |
% \bitheader{0-31} \\ |
| 326 |
% \begin{rightwordgroup}{RTP \\ Header} |
| 327 |
% \bitbox{2}{V=2} & \bitbox{1}{P} & \bitbox{1}{X} |
| 328 |
% & \bitbox{4}{CC} & \bitbox{1}{M} & \bitbox{7}{PT} |
| 329 |
% & \bitbox{16}{sequence number} \\ |
| 330 |
% \bitbox{32}{timestamp} |
| 331 |
% \end{rightwordgroup} \\ |
| 332 |
% \bitbox{32}{synchronization source (SSRC) identifier} \\ |
| 333 |
% \wordbox[tlr]{1}{contributing source (CSRC) identifiers} \\ |
| 334 |
% \wordbox[blr]{1}{$\cdots$} \\ |
| 335 |
% \begin{rightwordgroup}{RTP \\ Payload} |
| 336 |
% \wordbox[tlr]{3}{MPEG-4 Visual stream (byte aligned)} \\ |
| 337 |
% \bitbox[blr]{16}{} |
| 338 |
% & \bitbox{16}{\dots\emph{optional} RTP padding} |
| 339 |
% \end{rightwordgroup} |
| 340 |
% \end{bytefield} |
| 341 |
% \end{verbatim} |
| 342 |
% |
| 343 |
% Figure~\ref{fig:rtp-mpeg4} presents the typeset output of the |
| 344 |
% preceding code. Sections~\ref{sec:basic-cmds} and~\ref{sec:options} |
| 345 |
% explain each of the environments, macros, and arguments that were |
| 346 |
% utilized plus many additional features of the \pkgname{bytefield} |
| 347 |
% package. |
| 348 |
% |
| 349 |
% \begin{figure}[hbp] |
| 350 |
% \centering |
| 351 |
% \begin{lrbox}{\protocoldiagram} |
| 352 |
% \begin{bytefield}[bitwidth=1.1em]{32} |
| 353 |
% \bitheader{0-31} \\ |
| 354 |
% \begin{rightwordgroup}{RTP \\ Header} |
| 355 |
% \bitbox{2}{V=2} & \bitbox{1}{P} & \bitbox{1}{X} |
| 356 |
% & \bitbox{4}{CC} & \bitbox{1}{M} & \bitbox{7}{PT} |
| 357 |
% & \bitbox{16}{sequence number} \\ |
| 358 |
% \bitbox{32}{timestamp} |
| 359 |
% \end{rightwordgroup} \\ |
| 360 |
% \bitbox{32}{synchronization source (SSRC) identifier} \\ |
| 361 |
% \wordbox[tlr]{1}{contributing source (CSRC) identifiers} \\ |
| 362 |
% \wordbox[blr]{1}{$\cdots$} \\ |
| 363 |
% \begin{rightwordgroup}{RTP \\ Payload} |
| 364 |
% \wordbox[tlr]{3}{MPEG-4 Visual stream (byte aligned)} \\ |
| 365 |
% \bitbox[blr]{16}{} |
| 366 |
% & \bitbox{16}{\dots\emph{optional} RTP padding} |
| 367 |
% \end{rightwordgroup} |
| 368 |
% \end{bytefield} |
| 369 |
% \end{lrbox} |
| 370 |
% \makebox[0pt][c]{\usebox{\protocoldiagram}} |
| 371 |
% \caption{Sample \pkgname{bytefield} output} |
| 372 |
% \label{fig:rtp-mpeg4} |
| 373 |
% \end{figure} |
| 374 |
% |
| 375 |
% |
| 376 |
% \subsection{Basic commands} |
| 377 |
% \label{sec:basic-cmds} |
| 378 |
% |
| 379 |
% This section explains how to use the \pkgname{bytefield} package. It |
| 380 |
% lists all of the exported macros and environments in approximately |
| 381 |
% decreasing order of usefulness. |
| 382 |
% |
| 383 |
% \begin{decl} |
| 384 |
% \SpecialEnvIndex{bytefield} |
| 385 |
% |\begin{bytefield}| \oarg{options} \marg{bit-width} \\ |
| 386 |
% \meta{fields} \\ |
| 387 |
% |\end{bytefield}| |
| 388 |
% \end{decl} |
| 389 |
% |
| 390 |
% The \pkgname{bytefield} package's top-level environment is called, not |
| 391 |
% surprisingly, ``|bytefield|''. It takes one mandatory argument, which |
| 392 |
% is the number of bits in each word, and one optional argument, which |
| 393 |
% is a comma-separated list of \meta{key}=\meta{value} pairs, described |
| 394 |
% in Section~\ref{sec:options}, for formatting the bit-field's layout. |
| 395 |
% One can think of a |bytefield| as being analogous to a |tabular|: |
| 396 |
% words are separated by~``|\\|'', and fields within a word are |
| 397 |
% separated by~``|&|''. As in a |tabular|, ``|\\|'' accepts a |
| 398 |
% \meta{length} as an optional argument, and this specifies the amount |
| 399 |
% of additional vertical whitespace to include after the current word is |
| 400 |
% typeset. |
| 401 |
% |
| 402 |
% \begin{decl} |
| 403 |
% \SpecialUsageIndex{\bitbox} |
| 404 |
% |\bitbox| \oarg{sides} \marg{width} \oarg{options} \marg{text} \\ |
| 405 |
% \SpecialUsageIndex{\wordbox} |
| 406 |
% |\wordbox| \oarg{sides} \marg{height} \oarg{options} \marg{text} |
| 407 |
% \end{decl} |
| 408 |
% |
| 409 |
% The two main commands one uses within a \pkgname{bytefield} |
| 410 |
% environment are |\bitbox| and |\wordbox|. The former typesets a field |
| 411 |
% that is one or more bits wide and a single word tall. The latter |
| 412 |
% typesets a field that is an entire word wide and one or more words |
| 413 |
% tall. |
| 414 |
% |
| 415 |
% The first, optional, argument, \meta{sides}, is a list of letters |
| 416 |
% specifying which sides of the field box to draw---[|l|]eft, [|r|]ight, |
| 417 |
% [|t|]op, and/or [|b|]ottom. The default is ``|lrtb|'' (i.e.,~all |
| 418 |
% sides are drawn). The second, required, argument is the width in bits |
| 419 |
% of a bit box or the height in words of a word box. The third argument |
| 420 |
% is an optional, comma-separated list of \meta{key}=\meta{value} pairs, |
| 421 |
% described in Section~\ref{sec:options}. The fourth, required, |
| 422 |
% argument is the text to typeset within the box. It is typeset |
| 423 |
% horizontally centered within a vertically centered |\parbox|. Hence, |
| 424 |
% words will wrap, and ``|\\|'' can be used to break lines manually. |
| 425 |
% |
| 426 |
% The following example shows how to produce a simple 16-bit-wide field: |
| 427 |
% |
| 428 |
% \begin{verbatim} |
| 429 |
% \begin{bytefield}{16} |
| 430 |
% \wordbox{1}{A 16-bit field} \\ |
| 431 |
% \bitbox{8}{8 bits} & \bitbox{8}{8 more bits} \\ |
| 432 |
% \wordbox{2}{A 32-bit field. Note that text wraps within the box.} |
| 433 |
% \end{bytefield} |
| 434 |
% \end{verbatim} |
| 435 |
% |
| 436 |
% The resulting bit field looks like this: |
| 437 |
% |
| 438 |
% \begin{bffigure} |
| 439 |
% \begin{bytefield}{16} |
| 440 |
% \wordbox{1}{A 16-bit field} \\ |
| 441 |
% \bitbox{8}{8 bits} & \bitbox{8}{8 more bits} \\ |
| 442 |
% \wordbox{2}{A 32-bit field. Note that text wraps within the box.} |
| 443 |
% \end{bytefield} |
| 444 |
% \end{bffigure} |
| 445 |
% |
| 446 |
% It is the user's responsibility to ensure that the total number of |
| 447 |
% bits in each row adds up to the number of bits in a single word (the |
| 448 |
% mandatory argument to the |bytefield| environment); |
| 449 |
% \pkgname{bytefield} does not currently check for under- or overruns. |
| 450 |
% |
| 451 |
% Here's an example of using the \optname{bgcolor} option to fill each |
| 452 |
% box with a different color: |
| 453 |
% |
| 454 |
% \begin{verbatim} |
| 455 |
% \definecolor{lightcyan}{rgb}{0.84,1,1} |
| 456 |
% \definecolor{lightgreen}{rgb}{0.64,1,0.71} |
| 457 |
% \definecolor{lightred}{rgb}{1,0.7,0.71} |
| 458 |
% \begin{bytefield}[bitheight=\widthof{~Sign~}, |
| 459 |
% boxformatting={\centering\small}]{32} |
| 460 |
% \bitheader[endianness=big]{31,23,0} \\ |
| 461 |
% \bitbox{1}[bgcolor=lightcyan]{\rotatebox{90}{Sign}} & |
| 462 |
% \bitbox{8}[bgcolor=lightgreen]{Exponent} & |
| 463 |
% \bitbox{23}[bgcolor=lightred]{Mantissa} |
| 464 |
% \end{bytefield} |
| 465 |
% \end{verbatim} |
| 466 |
% |
| 467 |
% \begin{bffigure} |
| 468 |
% \definecolor{lightcyan}{rgb}{0.84,1,1} |
| 469 |
% \definecolor{lightgreen}{rgb}{0.64,1,0.71} |
| 470 |
% \definecolor{lightred}{rgb}{1,0.7,0.71} |
| 471 |
% \begin{bytefield}[bitheight=\widthof{~Sign~}, |
| 472 |
% boxformatting={\centering\small}]{32} |
| 473 |
% \bitheader[endianness=big]{31,23,0} \\ |
| 474 |
% \bitbox{1}[bgcolor=lightcyan]{\rotatebox{90}{Sign}} & |
| 475 |
% \bitbox{8}[bgcolor=lightgreen]{Exponent} & |
| 476 |
% \bitbox{23}[bgcolor=lightred]{Mantissa} |
| 477 |
% \end{bytefield} |
| 478 |
% \end{bffigure} |
| 479 |
% |
| 480 |
% Within a |\bitbox| or |\wordbox|, the \pkgname{bytefield} package |
| 481 |
% defines |\height|, |\depth|, |\totalheight|, and |\width| to the |
| 482 |
% corresponding dimensions of the box. Section~\ref{sec:tricks} gives an |
| 483 |
% example of how these lengths may be utilized. |
| 484 |
% |
| 485 |
% \begin{decl} |
| 486 |
% \SpecialUsageIndex{\bitboxes} |
| 487 |
% |\bitboxes| \oarg{sides} \marg{width} \oarg{options} \marg{tokens} \\ |
| 488 |
% |\bitboxes*| \oarg{sides} \marg{width} \oarg{options} \marg{tokens} \\ |
| 489 |
% \end{decl} |
| 490 |
% |
| 491 |
% The |\bitboxes| command provides a shortcut for typesetting a sequence |
| 492 |
% of fields of the same width. It takes essentially the same arguments |
| 493 |
% as |\bitbox| but interpets these differently. Instead of representing |
| 494 |
% a single piece of text to typeset within a field of width |
| 495 |
% \meta{width}, |\bitboxes|'s \meta{tokens} argument represents a list |
| 496 |
% of tokens (e.g,~individual characters), each of which is typeset |
| 497 |
% within a separate box of width \meta{width}. Consider, for example, |
| 498 |
% the following sequence of |\bitbox| commands: |
| 499 |
% |
| 500 |
% \begin{verbatim} |
| 501 |
% \begin{bytefield}{8} |
| 502 |
% \bitbox{1}{D} & \bitbox{1}{R} & \bitbox{1}{M} & \bitbox{1}{F} & |
| 503 |
% \bitbox{1}{S} & \bitbox{1}{L} & \bitbox{1}{T} & \bitbox{1}{D} |
| 504 |
% \end{bytefield} |
| 505 |
% \end{verbatim} |
| 506 |
% |
| 507 |
% \begin{bffigure} |
| 508 |
% \begin{bytefield}{8} |
| 509 |
% \bitbox{1}{D} & \bitbox{1}{R} & \bitbox{1}{M} & \bitbox{1}{F} & |
| 510 |
% \bitbox{1}{S} & \bitbox{1}{L} & \bitbox{1}{T} & \bitbox{1}{D} |
| 511 |
% \end{bytefield} |
| 512 |
% \end{bffigure} |
| 513 |
% |
| 514 |
% \noindent |
| 515 |
% With |\bitboxes| this can be abbreviated to |
| 516 |
% |
| 517 |
% \begin{verbatim} |
| 518 |
% \begin{bytefield}{8} |
| 519 |
% \bitboxes{1}{DRMFSLTD} |
| 520 |
% \end{bytefield} |
| 521 |
% \end{verbatim} |
| 522 |
% |
| 523 |
% Spaces are ignored within |\bitboxes|'s \meta{text} argument, and |
| 524 |
% curly braces can be used to group multiple characters into a single |
| 525 |
% token: |
| 526 |
% |
| 527 |
% \begin{verbatim} |
| 528 |
% \begin{bytefield}{24} |
| 529 |
% \bitboxes{3}{{DO} {RE} {MI} {FA} {SOL} {LA} {TI} {DO}} |
| 530 |
% \end{bytefield} |
| 531 |
% \end{verbatim} |
| 532 |
% |
| 533 |
% \begin{bffigure} |
| 534 |
% \begin{bytefield}{24} |
| 535 |
% \bitboxes{3}{{DO} {RE} {MI} {FA} {SOL} {LA} {TI} {DO}} |
| 536 |
% \end{bytefield} |
| 537 |
% \end{bffigure} |
| 538 |
% |
| 539 |
% The starred form of |\bitboxes| is identical except that it suppresses |
| 540 |
% all internal vertical lines. It can therefore be quite convenient for |
| 541 |
% typesetting binary constants: |
| 542 |
% |
| 543 |
% \begin{verbatim} |
| 544 |
% \begin{bytefield}{16} |
| 545 |
% \bitboxes*{1}{01000010} & \bitbox{4}{src\strut} & |
| 546 |
% \bitbox{4}{dest\strut} & \bitbox{4}{const\strut} |
| 547 |
% \end{bytefield} |
| 548 |
% \end{verbatim} |
| 549 |
% |
| 550 |
% \begin{bffigure} |
| 551 |
% \begin{bytefield}{16} |
| 552 |
% \bitboxes*{1}{01000010} & \bitbox{4}{src\strut} & |
| 553 |
% \bitbox{4}{dest\strut} & \bitbox{4}{const\strut} |
| 554 |
% \end{bytefield} |
| 555 |
% \end{bffigure} |
| 556 |
% |
| 557 |
% \begin{decl} |
| 558 |
% \SpecialUsageIndex{\bitheader} |
| 559 |
% |\bitheader| \oarg{options} \marg{bit-positions} |
| 560 |
% \end{decl} |
| 561 |
% |
| 562 |
% To make the bit field more readable, it helps to label bit positions |
| 563 |
% across the top. The |\bitheader| command provides a flexible way to |
| 564 |
% do that. The optional argument is a comma-separated list of |
| 565 |
% \meta{key}=\meta{value} pairs from the set described in |
| 566 |
% Section~\ref{sec:options}. In practice, the only parameters that are |
| 567 |
% meaningful in the context of |\bitheader| are |bitformatting|, |
| 568 |
% |endianness|, and |lsb|. See Section~\ref{sec:options} for |
| 569 |
% descriptions and examples of those parameters. |
| 570 |
% |
| 571 |
% |\bitheader|'s mandatory argument, \meta{bit-positions}, is a |
| 572 |
% comma-separated list of bit positions to label. For example, |
| 573 |
% ``|0,2,4,6,8,10,12,14|'' means to label those bit positions. The |
| 574 |
% numbers must be listed in increasing order. (Use the |
| 575 |
% \optname{endianness} parameter to display the header in reverse |
| 576 |
% order.) Hyphen-separated ranges are also valid. For example, |
| 577 |
% ``|0-15|'' means to label all bits from~0 to~15, inclusive. Ranges |
| 578 |
% and single numbers can even be intermixed, as in ``|0-3,8,12-15|''. |
| 579 |
% |
| 580 |
% The following example shows how |\bitheader| may be used: |
| 581 |
% |
| 582 |
% \begin{verbatim} |
| 583 |
% \begin{bytefield}{32} |
| 584 |
% \bitheader{0-31} \\ |
| 585 |
% \bitbox{4}{Four} & \bitbox{8}{Eight} & |
| 586 |
% \bitbox{16}{Sixteen} & \bitbox{4}{Four} |
| 587 |
% \end{bytefield} |
| 588 |
% \end{verbatim} |
| 589 |
% |
| 590 |
% The resulting bit field looks like this: |
| 591 |
% |
| 592 |
% \begin{bffigure} |
| 593 |
% \begin{bytefield}{32} |
| 594 |
% \bitheader{0-31} \\ |
| 595 |
% \bitbox{4}{Four} & \bitbox{8}{Eight} & |
| 596 |
% \bitbox{16}{Sixteen} & \bitbox{4}{Four} |
| 597 |
% \end{bytefield} |
| 598 |
% \end{bffigure} |
| 599 |
% |
| 600 |
% |
| 601 |
% \begin{decl} |
| 602 |
% \SpecialEnvIndex{rightwordgroup} |
| 603 |
% |\begin{rightwordgroup}| \oarg{options} \marg{text} \\ |
| 604 |
% \meta{rows of bit boxes and word boxes} \\ |
| 605 |
% |\end{rightwordgroup}| \\ |
| 606 |
% \\ |
| 607 |
% \SpecialEnvIndex{leftwordgroup} |
| 608 |
% |\begin{leftwordgroup}| \oarg{options} \marg{text} \\ |
| 609 |
% \meta{rows of bit boxes and word boxes} \\ |
| 610 |
% |\end{leftwordgroup}| |
| 611 |
% \end{decl} |
| 612 |
% |
| 613 |
% When a set of words functions as a single, logical unit, it helps to |
| 614 |
% group these words together visually. All words defined between |
| 615 |
% |\begin{rightwordgroup}| and |\end{rightwordgroup}| will be labeled on |
| 616 |
% the right with \meta{text}. Similarly, all words defined between |
| 617 |
% |\begin{leftwordgroup}| and |\end{leftwordgroup}| will be labeled on |
| 618 |
% the left with \meta{text}. |\begin{|\meta{side}|wordgroup}| must lie |
| 619 |
% at the beginning of a row (i.e.,~right after a ``|\\|''), and |
| 620 |
% |\end{|\meta{side}|wordgroup}| must lie right \emph{before} the end of |
| 621 |
% the row (i.e.,~right before a ``|\\|''). |
| 622 |
% |
| 623 |
% The optional argument is a comma-separated list of |
| 624 |
% \meta{key}=\meta{value} pairs from the set described in |
| 625 |
% Section~\ref{sec:options}. In practice, only \optname{curlystyle}, |
| 626 |
% \optname{leftcurlystyle}, and \optname{rightcurlystyle}, make sense |
| 627 |
% within the context of a |\begin{|\meta{side}|wordgroup}|. |
| 628 |
% |
| 629 |
% Unlike other \LaTeX\ environments, |rightwordgroup| and |
| 630 |
% |leftwordgroup| do not have to nest properly with each other. |
| 631 |
% However, they cannot overlap themselves. In other words, |
| 632 |
% |\begin{rightwordgroup}|\ellipsis |\begin{leftwordgroup}|\ellipsis |
| 633 |
% |\end{rightwordgroup}|\ellipsis |\end{leftwordgroup}| is a valid |
| 634 |
% sequence, but |\begin{rightwordgroup}|\ellipsis |
| 635 |
% |\begin{rightwordgroup}|\ellipsis |\end{rightwordgroup}|\ellipsis |
| 636 |
% |\end{rightwordgroup}| is not. |
| 637 |
% |
| 638 |
% The following example presents the basic usage of |
| 639 |
% |\begin{rightwordgroup}| and |\end{rightwordgroup}|: |
| 640 |
% |
| 641 |
% \begin{verbatim} |
| 642 |
% \begin{bytefield}{16} |
| 643 |
% \bitheader{0,7,8,15} \\ |
| 644 |
% \begin{rightwordgroup}{Header} |
| 645 |
% \bitbox{4}{Tag} & \bitbox{12}{Mask} \\ |
| 646 |
% \bitbox{8}{Source} & \bitbox{8}{Destination} |
| 647 |
% \end{rightwordgroup} \\ |
| 648 |
% \wordbox{3}{Data} |
| 649 |
% \end{bytefield} |
| 650 |
% \end{verbatim} |
| 651 |
% |
| 652 |
% \noindent |
| 653 |
% Note the juxtaposition of ``|\\|'' to the |\begin{rightwordgroup}| and |
| 654 |
% the |\end{rightwordgroup}| in the above. The resulting bit field |
| 655 |
% looks like this: |
| 656 |
% |
| 657 |
% \begin{bffigure} |
| 658 |
% \begin{bytefield}{16} |
| 659 |
% \bitheader{0,7,8,15} \\ |
| 660 |
% \begin{rightwordgroup}{Header} |
| 661 |
% \bitbox{4}{Tag} & \bitbox{12}{Mask} \\ |
| 662 |
% \bitbox{8}{Source} & \bitbox{8}{Destination} |
| 663 |
% \end{rightwordgroup} \\ |
| 664 |
% \wordbox{3}{Data} |
| 665 |
% \end{bytefield} |
| 666 |
% \end{bffigure} |
| 667 |
% |
| 668 |
% \noindent |
| 669 |
% As a more complex example, the following nests left and right labels: |
| 670 |
% |
| 671 |
% \begin{verbatim} |
| 672 |
% \begin{bytefield}{16} |
| 673 |
% \bitheader{0,7,8,15} \\ |
| 674 |
% \begin{rightwordgroup}{Header} |
| 675 |
% \bitbox{4}{Tag} & \bitbox{12}{Mask} \\ |
| 676 |
% \begin{leftwordgroup}{Node IDs} |
| 677 |
% \bitbox{8}{Source} & \bitbox{8}{Destination} |
| 678 |
% \end{leftwordgroup} |
| 679 |
% \end{rightwordgroup} \\ |
| 680 |
% \wordbox{3}{Data} |
| 681 |
% \end{bytefield} |
| 682 |
% \end{verbatim} |
| 683 |
% |
| 684 |
% \begin{bffigure} |
| 685 |
% \begin{bytefield}{16} |
| 686 |
% \bitheader{0,7,8,15} \\ |
| 687 |
% \begin{rightwordgroup}{Header} |
| 688 |
% \bitbox{4}{Tag} & \bitbox{12}{Mask} \\ |
| 689 |
% \begin{leftwordgroup}{Node IDs} |
| 690 |
% \bitbox{8}{Source} & \bitbox{8}{Destination} |
| 691 |
% \end{leftwordgroup} |
| 692 |
% \end{rightwordgroup} \\ |
| 693 |
% \wordbox{3}{Data} |
| 694 |
% \end{bytefield} |
| 695 |
% \end{bffigure} |
| 696 |
% |
| 697 |
% \noindent |
| 698 |
% Because |rightwordgroup| and |leftwordgroup| are not required to nest |
| 699 |
% properly, the resulting bit field would look the same if the |
| 700 |
% |\end{leftwordgroup}| and |\end{rightwordgroup}| were swapped. Again, |
| 701 |
% note the justaposition of ``|\\|'' to the various word-grouping |
| 702 |
% commands in the above. |
| 703 |
% |
| 704 |
% \begin{decl} |
| 705 |
% \SpecialUsageIndex{\skippedwords} |
| 706 |
% |\skippedwords| |
| 707 |
% \end{decl} |
| 708 |
% |
| 709 |
% Draw a graphic representing a number of words that are not shown. |
| 710 |
% |\skippedwords| is intended to work with the \meta{sides} argument |
| 711 |
% to |\wordbox|, as in the following example: |
| 712 |
% |
| 713 |
% \begin{verbatim} |
| 714 |
% \begin{bytefield}{16} |
| 715 |
% \wordbox{1}{Some data} \\ |
| 716 |
% \wordbox[lrt]{1}{Lots of data} \\ |
| 717 |
% \skippedwords \\ |
| 718 |
% \wordbox[lrb]{1}{} \\ |
| 719 |
% \wordbox{2}{More data} |
| 720 |
% \end{bytefield} |
| 721 |
% \end{verbatim} |
| 722 |
% |
| 723 |
% \begin{bffigure} |
| 724 |
% \begin{bytefield}{16} |
| 725 |
% \wordbox{1}{Some data} \\ |
| 726 |
% \wordbox[lrt]{1}{Lots of data} \\ |
| 727 |
% \skippedwords \\ |
| 728 |
% \wordbox[lrb]{1}{} \\ |
| 729 |
% \wordbox{2}{More data} |
| 730 |
% \end{bytefield} |
| 731 |
% \end{bffigure} |
| 732 |
% |
| 733 |
% \begin{decl} |
| 734 |
% \SpecialUsageIndex{\bytefieldsetup} |
| 735 |
% |\bytefieldsetup| \marg{options} |
| 736 |
% \end{decl} |
| 737 |
% |
| 738 |
% Alter the formatting of all subsequent bit fields. |
| 739 |
% Section~\ref{sec:options} describes the possible values for each |
| 740 |
% \meta{key}=\meta{value} pair in the comma-separated list that |
| 741 |
% |\bytefieldsetup| accepts as its argument. Note that changes made |
| 742 |
% with |\bytefieldsetup| are local to their current scope. Hence, if |
| 743 |
% used within an environment (e.g.,~|figure|), |\bytefieldsetup| does |
| 744 |
% not impact bit fields drawn outside that environment. |
| 745 |
% |
| 746 |
% |
| 747 |
% \subsection{Formatting options} |
| 748 |
% \label{sec:options} |
| 749 |
% |
| 750 |
% A document author can customize many of the \pkgname{bytefield} |
| 751 |
% package's figure-formatting parameters, either globally or on a |
| 752 |
% per-figure basis. The parameters described below can be specified in |
| 753 |
% six locations: |
| 754 |
% |
| 755 |
% \begin{itemize} |
| 756 |
% \item as package options (i.e.,~in the |
| 757 |
% |\usepackage|\oarg{options}|{bytefield}| line), which affects all |
| 758 |
% |bytefield| environments in the entire document, |
| 759 |
% |
| 760 |
% \item anywhere in the document using the |\bytefieldsetup| command, |
| 761 |
% which affects all subsequent |bytefield| environments in the |
| 762 |
% current scope, |
| 763 |
% |
| 764 |
% \item as the optional argument to a |\begin{bytefield}|, which |
| 765 |
% affects only that single bit-field figure, or |
| 766 |
% |
| 767 |
% \item as the optional argument to a |\bitheader|, which affects only |
| 768 |
% that particular header. (Only a few parameters are meaningful in |
| 769 |
% this context.) |
| 770 |
% |
| 771 |
% \item as the optional argument to a |\begin{leftwordgroup}| or |
| 772 |
% |\begin{rightwordgroup}|, which affects only that particular word |
| 773 |
% group. (Only a few parameters are meaningful in this context.) |
| 774 |
% |
| 775 |
% \item as the second optional argument to a \cs{bitbox}, \cs{wordbox}, |
| 776 |
% or \cs{bitboxes}, which affects only that particular box. (Only a |
| 777 |
% few parameters are meaningful in this context.) |
| 778 |
% \end{itemize} |
| 779 |
% |
| 780 |
% Unfortunately, \LaTeX\ tends to abort with a ``\texttt{TeX capacity |
| 781 |
% exceeded}'' or ``\texttt{Missing \cs{endcsname} inserted}'' error |
| 782 |
% when a control sequence (i.e.,~|\|\meta{name}) or |\|\meta{symbol}) is |
| 783 |
% encountered within the optional argument to |\usepackage|. Hence, |
| 784 |
% parameters that typically expect a control sequence in their |
| 785 |
% argument---in particular, \optname{bitformatting}, |
| 786 |
% \optname{boxformatting}, \optname{leftcurly}, and |
| 787 |
% \optname{rightcurly}---should best be avoided within the |
| 788 |
% |\usepackage|\oarg{options}|{bytefield}| line. |
| 789 |
% |
| 790 |
% \begin{decl} |
| 791 |
% \optname{bitwidth} = \meta{length} \\ |
| 792 |
% \optname{bitheight} = \meta{length} |
| 793 |
% \end{decl} |
| 794 |
% |
| 795 |
% The above parameters represent the width and height of each bit in a |
| 796 |
% bit field. The default value of \optname{bitwidth} is the width of |
| 797 |
% ``|{\tiny 99i}|'', i.e.,~the width of a two-digit number plus a small |
| 798 |
% amount of extra space. This enables |\bitheader| to show two-digit |
| 799 |
% numbers without overlap. The default value of \optname{bitheight} |
| 800 |
% is~|2ex|, which should allow a normal piece of text to appear within a |
| 801 |
% |\bitbox| or |\wordbox| without abutting the box's top or bottom edge. |
| 802 |
% |
| 803 |
% As a special case, if \optname{bitwidth} is set to the word |
| 804 |
% ``|auto|'', it will be set to the width of ``|99i|'' in the current |
| 805 |
% bit-number formatting (see \optname{bitformatting} below). This |
| 806 |
% feature provides a convenient way to adjust the bit width after a |
| 807 |
% formatting change. |
| 808 |
% |
| 809 |
% \begin{decl} |
| 810 |
% \optname{endianness} = |little| \emph{or} |big| |
| 811 |
% \end{decl} |
| 812 |
% |
| 813 |
% Specify either little-endian (left-to-right) or big-endian |
| 814 |
% (right-to-left) ordering of the bit numbers. The default is |
| 815 |
% little-endian numbering. Contrast the following two examples. The |
| 816 |
% first formats a bit field in little-endian ordering using an explicit |
| 817 |
% |endianness=little|, and the second formats the same bit field in |
| 818 |
% big-endian ordering using |endianness=big|. |
| 819 |
% |
| 820 |
% \begin{verbatim} |
| 821 |
% \begin{bytefield}[endianness=little,bitwidth=0.11111\linewidth]{8} |
| 822 |
% \bitheader{0-7} \\ |
| 823 |
% \bitbox{1}{Res} & \bitbox{1}{BE} & \bitbox{1}{CF} |
| 824 |
% & \bitbox{3}{$\mbox{Name\_Len}-1$} & \bitbox{2}{Len\_Len} \\ |
| 825 |
% \end{bytefield} |
| 826 |
% \end{verbatim} |
| 827 |
% |
| 828 |
% \begin{bffigure} |
| 829 |
% \begin{bytefield}[endianness=little,bitwidth=0.11111\linewidth]{8} |
| 830 |
% \bitheader{0-7} \\ |
| 831 |
% \bitbox{1}{Res} & \bitbox{1}{BE} & \bitbox{1}{CF} |
| 832 |
% & \bitbox{3}{$\mbox{Name\_Len}-1$} & \bitbox{2}{Len\_Len} \\ |
| 833 |
% \end{bytefield} |
| 834 |
% \end{bffigure} |
| 835 |
% |
| 836 |
% \begin{verbatim} |
| 837 |
% \begin{bytefield}[endianness=big,bitwidth=0.11111\linewidth]{8} |
| 838 |
% \bitheader{0-7} \\ |
| 839 |
% \bitbox{2}{Len\_Len} & \bitbox{3}{$\mbox{Name\_Len}-1$} |
| 840 |
% & \bitbox{1}{CF} & \bitbox{1}{BE} & \bitbox{1}{Res} \\ |
| 841 |
% \end{bytefield} |
| 842 |
% \end{verbatim} |
| 843 |
% |
| 844 |
% \begin{bffigure} |
| 845 |
% \begin{bytefield}[endianness=big,bitwidth=0.11111\linewidth]{8} |
| 846 |
% \bitheader{0-7} \\ |
| 847 |
% \bitbox{2}{Len\_Len} & \bitbox{3}{$\mbox{Name\_Len}-1$} |
| 848 |
% & \bitbox{1}{CF} & \bitbox{1}{BE} & \bitbox{1}{Res} \\ |
| 849 |
% \end{bytefield} |
| 850 |
% \end{bffigure} |
| 851 |
% |
| 852 |
% \begin{decl} |
| 853 |
% \optname{bitformatting} = \meta{command} \emph{or} \marg{commands} |
| 854 |
% \end{decl} |
| 855 |
% |
| 856 |
% The numbers that appear in a bit header are typeset in the |
| 857 |
% \optname{bitformatting} style, which defaults to |\tiny|. To alter |
| 858 |
% the style of bit numbers in the bit header, set |
| 859 |
% \optname{bitformatting} to a macro that takes a single argument (like |
| 860 |
% |\textbf|) or no arguments (like |\small|). Groups of commands |
| 861 |
% (e.g.,~|{\large\itshape}|) are also acceptable. |
| 862 |
% |
| 863 |
% When \optname{bitformatting} is set, \optname{bitwidth} usually needs |
| 864 |
% to be recalculated as well to ensure that a correct amount of spacing |
| 865 |
% surrounds each number in the bit header. As described above, setting |
| 866 |
% |bitwidth=auto| is a convenient shortcut for recalculating the |
| 867 |
% bit-width in the common case of bit fields containing no more than 99 |
| 868 |
% bits per line and no particularly wide labels in bit boxes that |
| 869 |
% contain only a few bits. |
| 870 |
% |
| 871 |
% The following example shows how to use \optname{bitformatting} and |
| 872 |
% \optname{bitwidth} to format a bit header with small, boldface text: |
| 873 |
% |
| 874 |
% \begin{verbatim} |
| 875 |
% \begin{bytefield}[bitformatting={\small\bfseries}, |
| 876 |
% bitwidth=auto, |
| 877 |
% endianness=big]{20} |
| 878 |
% \bitheader{0-19} \\ |
| 879 |
% \bitbox{1}{\tiny F/E} & \bitbox{1}{\tiny T0} & \bitbox{1}{\tiny T1} |
| 880 |
% & \bitbox{1}{\tiny Fwd} & \bitbox{16}{Data value} \\ |
| 881 |
% \end{bytefield} |
| 882 |
% \end{verbatim} |
| 883 |
% |
| 884 |
% \noindent |
| 885 |
% The resulting bit field looks like this: |
| 886 |
% |
| 887 |
% \begin{bffigure} |
| 888 |
% \begin{bytefield}[bitformatting={\small\bfseries}, |
| 889 |
% bitwidth=auto, |
| 890 |
% endianness=big]{20} |
| 891 |
% \bitheader{0-19} \\ |
| 892 |
% \bitbox{1}{\tiny F/E} & \bitbox{1}{\tiny T0} & \bitbox{1}{\tiny T1} |
| 893 |
% & \bitbox{1}{\tiny Fwd} & \bitbox{16}{Data value} \\ |
| 894 |
% \end{bytefield} |
| 895 |
% \end{bffigure} |
| 896 |
% |
| 897 |
% \begin{decl} |
| 898 |
% \optname{boxformatting} = \meta{command} \emph{or} \marg{commands} |
| 899 |
% \end{decl} |
| 900 |
% |
| 901 |
% The text that appears in a |\bitbox| or |\wordbox| is formatted in the |
| 902 |
% \optname{boxformatting} style, which defaults to |\centering|. To |
| 903 |
% alter the style of bit numbers in the bit header, set |
| 904 |
% \optname{boxformatting} to a macro that takes a single argument (like |
| 905 |
% |\textbf| but not |\textbf|---see below) or no arguments (like |
| 906 |
% |\small|). Groups of commands (e.g.,~|{\large\itshape}|) are also |
| 907 |
% acceptable. |
| 908 |
% |
| 909 |
% If \optname{boxformatting} is set to a macro that takes an argument, |
| 910 |
% the macro must be defined as a ``long'' macro, which means it can |
| 911 |
% accept more than one paragraph as an argument. Commands defined with |
| 912 |
% |\newcommand| are automatically made long, but commands defined with |
| 913 |
% |\newcommand*| are not. \LaTeX's |\text|\dots\ formatting commands |
| 914 |
% (e.g.,~|\textbf|) are not long and therefore cannot be used directly |
| 915 |
% in \optname{boxformatting}; use the zero-argument versions |
| 916 |
% (e.g.,~|\bfseries|) instead. |
| 917 |
% |
| 918 |
% The following example shows how to use \optname{boxformatting} to |
| 919 |
% format the text within each box horizontally centered and italicized: |
| 920 |
% |
| 921 |
% \begin{verbatim} |
| 922 |
% \begin{bytefield}[boxformatting={\centering\itshape}, |
| 923 |
% bitwidth=1.5em, |
| 924 |
% endianness=big]{20} |
| 925 |
% \bitheader{0-19} \\ |
| 926 |
% \bitbox{1}{\tiny F/E} & \bitbox{1}{\tiny T0} & \bitbox{1}{\tiny T1} |
| 927 |
% & \bitbox{1}{\tiny Fwd} & \bitbox{16}{Data value} \\ |
| 928 |
% \end{bytefield} |
| 929 |
% \end{verbatim} |
| 930 |
% |
| 931 |
% \noindent |
| 932 |
% The resulting bit field looks like this: |
| 933 |
% |
| 934 |
% \begin{bffigure} |
| 935 |
% \begin{bytefield}[boxformatting={\centering\itshape}, |
| 936 |
% bitwidth=1.5em, |
| 937 |
% endianness=big]{20} |
| 938 |
% \bitheader{0-19} \\ |
| 939 |
% \bitbox{1}{\tiny F/E} & \bitbox{1}{\tiny T0} & \bitbox{1}{\tiny T1} |
| 940 |
% & \bitbox{1}{\tiny Fwd} & \bitbox{16}{Data value} \\ |
| 941 |
% \end{bytefield} |
| 942 |
% \end{bffigure} |
| 943 |
% |
| 944 |
% \begin{decl} |
| 945 |
% \optname{bgcolor} = \meta{color} |
| 946 |
% \end{decl} |
| 947 |
% |
| 948 |
% Bit and word boxes are normally left unfilled. The \optname{bgcolor} |
| 949 |
% option fills them with a specified background color. A document will |
| 950 |
% need to include the \pkgname{color}, \pkgname{xcolor}, or similar |
| 951 |
% package to expose color names to \pkgname{bytefield}. The |
| 952 |
% \optname{boxformatting} option described above can be used to set the |
| 953 |
% foreground color. |
| 954 |
% |
| 955 |
% \begin{decl} |
| 956 |
% \optname{leftcurly} = \meta{delimiter} \\ |
| 957 |
% \optname{rightcurly} = \meta{delimiter} |
| 958 |
% \end{decl} |
| 959 |
% |
| 960 |
% Word groups are normally indicated by a curly brace spanning all of |
| 961 |
% its rows. However, the curly brace can be replaced by any other |
| 962 |
% extensible math delimiter (i.e.,~a symbol that can meaningfully follow |
| 963 |
% |\left| or |\right| in math mode) via a suitable redefinition of |
| 964 |
% \optname{leftcurly} or \optname{rightcurly}. As in math mode, ``|.|'' |
| 965 |
% means ``no symbol'', as in the following example (courtesy of |
| 966 |
% Steven~R. King): |
| 967 |
% |
| 968 |
% \begin{verbatim} |
| 969 |
% \begin{bytefield}[rightcurly=., rightcurlyspace=0pt]{32} |
| 970 |
% \bitheader[endianness=big]{0,7,8,15,16,23,24,31} \\ |
| 971 |
% \begin{rightwordgroup}{0Ch} |
| 972 |
% \bitbox{8}{Byte 15 \\ \tiny (highest address)} |
| 973 |
% & \bitbox{8}{Byte 14} |
| 974 |
% & \bitbox{8}{Byte 13} |
| 975 |
% & \bitbox{8}{Byte 12} |
| 976 |
% \end{rightwordgroup} \\ |
| 977 |
% \begin{rightwordgroup}{08h} |
| 978 |
% \bitbox{32}{Long 0} |
| 979 |
% \end{rightwordgroup} \\ |
| 980 |
% \begin{rightwordgroup}{04h} |
| 981 |
% \bitbox{16}{Word 1} & \bitbox{16}{Word 0} |
| 982 |
% \end{rightwordgroup} \\ |
| 983 |
% \begin{rightwordgroup}{00h} |
| 984 |
% \bitbox{8}{Byte 3} |
| 985 |
% & \bitbox{8}{Byte 2} |
| 986 |
% & \bitbox{8}{Byte 1} |
| 987 |
% & \bitbox{8}{Byte 0 \\ \tiny (lowest address)} |
| 988 |
% \end{rightwordgroup} |
| 989 |
% \end{bytefield} |
| 990 |
% \end{verbatim} |
| 991 |
% |
| 992 |
% \begin{bffigure} |
| 993 |
% \begin{bytefield}[rightcurly=., rightcurlyspace=0pt]{32} |
| 994 |
% \bitheader[endianness=big]{0,7,8,15,16,23,24,31} \\ |
| 995 |
% \begin{rightwordgroup}{0Ch} |
| 996 |
% \bitbox{8}{Byte 15 \\ \tiny (highest address)} |
| 997 |
% & \bitbox{8}{Byte 14} |
| 998 |
% & \bitbox{8}{Byte 13} |
| 999 |
% & \bitbox{8}{Byte 12} |
| 1000 |
% \end{rightwordgroup} \\ |
| 1001 |
% \begin{rightwordgroup}{08h} |
| 1002 |
% \bitbox{32}{Long 0} |
| 1003 |
% \end{rightwordgroup} \\ |
| 1004 |
% \begin{rightwordgroup}{04h} |
| 1005 |
% \bitbox{16}{Word 1} & \bitbox{16}{Word 0} |
| 1006 |
% \end{rightwordgroup} \\ |
| 1007 |
% \begin{rightwordgroup}{00h} |
| 1008 |
% \bitbox{8}{Byte 3} |
| 1009 |
% & \bitbox{8}{Byte 2} |
| 1010 |
% & \bitbox{8}{Byte 1} |
| 1011 |
% & \bitbox{8}{Byte 0 \\ \tiny (lowest address)} |
| 1012 |
% \end{rightwordgroup} |
| 1013 |
% \end{bytefield} |
| 1014 |
% \end{bffigure} |
| 1015 |
% |
| 1016 |
% \begin{decl} |
| 1017 |
% \optname{leftcurlyspace} = \meta{length} \\ |
| 1018 |
% \optname{rightcurlyspace} = \meta{length} \\ |
| 1019 |
% \optname{curlyspace} = \meta{length} |
| 1020 |
% \end{decl} |
| 1021 |
% |
| 1022 |
% \optname{leftcurlyspace} and \optname{rightcurlyspace} specify the |
| 1023 |
% space to insert between the bit field and the curly brace in a left or |
| 1024 |
% right word group (default:~|1ex|). Setting \optname{curlyspace} is a |
| 1025 |
% shortcut for setting both \optname{leftcurlyspace} and |
| 1026 |
% \optname{rightcurlyspace} to the same value. |
| 1027 |
% |
| 1028 |
% \begin{decl} |
| 1029 |
% |leftlabelspace| = \meta{length} \\ |
| 1030 |
% |rightlabelspace| = \meta{length} \\ |
| 1031 |
% |labelspace| = \meta{length} |
| 1032 |
% \end{decl} |
| 1033 |
% |
| 1034 |
% |leftlabelspace| and |rightlabelspace| specify the space to insert |
| 1035 |
% between the curly brace and the text label in a left or right word |
| 1036 |
% group (default:~|0.5ex|). Setting |labelspace| is a shortcut for |
| 1037 |
% setting both |leftlabelbrace| and |rightlabelspace| to the same value. |
| 1038 |
% |
| 1039 |
% \bigskip |
| 1040 |
% |
| 1041 |
% Figure~\ref{fig:rightspace} illustrates the juxtaposition of |
| 1042 |
% \optname{rightcurlyspace} and |rightlabelspace| to a word group and |
| 1043 |
% its label. The \optname{leftcurlyspace} and |leftlabelspace| |
| 1044 |
% parameters are symmetric. |
| 1045 |
% |
| 1046 |
% \begin{figure}[htbp] |
| 1047 |
% \[ |
| 1048 |
% \begin{array}{*5{@{}c}@{}} |
| 1049 |
% \cline{1-1} |
| 1050 |
% \multicolumn{1}{c|}{\cdots\quad\meta{bit-field rows}} |
| 1051 |
% & |
| 1052 |
% & \left.\rule{0pt}{3ex}\right\} |
| 1053 |
% & |
| 1054 |
% & \meta{label} \\[2ex] \cline{1-1} |
| 1055 |
% & \underbrace{\hspace*{6em}}_{\mathtt{rightcurlyspace}} |
| 1056 |
% & |
| 1057 |
% & \underbrace{\hspace*{6em}}_{\mathtt{rightlabelspace}} \\ |
| 1058 |
% \end{array} |
| 1059 |
% \] |
| 1060 |
% \caption{Role of \texttt{rightcurlyspace} and \texttt{rightlabelspace}} |
| 1061 |
% \label{fig:rightspace} |
| 1062 |
% \end{figure} |
| 1063 |
% |
| 1064 |
% \begin{decl} |
| 1065 |
% \optname{leftcurlyshrinkage} = \meta{length} \\ |
| 1066 |
% \optname{rightcurlyshrinkage} = \meta{length} \\ |
| 1067 |
% \optname{curlyshrinkage} = \meta{length} |
| 1068 |
% \end{decl} |
| 1069 |
% |
| 1070 |
% In \TeX/\LaTeX, the height of a curly brace does not include the tips. |
| 1071 |
% Hence, in a word group label, the tips of the curly brace will extend |
| 1072 |
% beyond the height of the word group. |
| 1073 |
% \optname{leftcurlyshrinkage}\slash \optname{rightcurlyshrinkage} is an |
| 1074 |
% amount by which to reduce the height of the curly brace in a |
| 1075 |
% left\slash right word group's label. Setting \optname{curlyshrinkage} |
| 1076 |
% is a shortcut for setting both \optname{leftcurlyshrinkage} and |
| 1077 |
% \optname{rightcurlyshrinkage} to the same value. Shrinkages default |
| 1078 |
% to~|5pt|, and it is extremely unlikely that one would ever need to |
| 1079 |
% change them. Nevertheless, these parameters are included here in case |
| 1080 |
% a document is typeset with a math font containing radically different |
| 1081 |
% curly braces from the ones that come with \TeX/\LaTeX\ or that |
| 1082 |
% replaces the curly braces (using \optname{leftcurly}\slash |
| 1083 |
% \optname{rightcurly}, described above) with symbols of substantially |
| 1084 |
% different heights. |
| 1085 |
% |
| 1086 |
% \begin{decl} |
| 1087 |
% \optname{leftcurlystyle} = \meta{command} \\ |
| 1088 |
% \optname{rightcurlystyle} = \meta{command} \\ |
| 1089 |
% \optname{curlystyle} = \meta{command} |
| 1090 |
% \end{decl} |
| 1091 |
% |
| 1092 |
% Provide a macro that will be invoked before the code that draws left, |
| 1093 |
% right, or both curly braces. The macro must accept either zero or one |
| 1094 |
% argument. It can be used, for example, to color the curly brace: |
| 1095 |
% |
| 1096 |
% \begin{verbatim} |
| 1097 |
% \begin{bytefield}[curlystyle=\color{blue}]{16} |
| 1098 |
% \bitheader{0-15} \\ |
| 1099 |
% \begin{rightwordgroup}{Sign-extended} |
| 1100 |
% \bitbox{4}{Tag} & \bitbox{12}{Data} |
| 1101 |
% \end{rightwordgroup} \\ |
| 1102 |
% \end{bytefield} |
| 1103 |
% \end{verbatim} |
| 1104 |
% |
| 1105 |
% \begin{bffigure} |
| 1106 |
% \begin{bytefield}[curlystyle=\color{blue}]{16} |
| 1107 |
% \bitheader{0-15} \\ |
| 1108 |
% \begin{rightwordgroup}{Sign-extended} |
| 1109 |
% \bitbox{4}{Tag} & \bitbox{12}{Data} |
| 1110 |
% \end{rightwordgroup} \\ |
| 1111 |
% \end{bytefield} |
| 1112 |
% \end{bffigure} |
| 1113 |
% |
| 1114 |
% \noindent |
| 1115 |
% or |
| 1116 |
% |
| 1117 |
% \begin{verbatim} |
| 1118 |
% \begin{bytefield}{16} |
| 1119 |
% \bitheader{0-15} \\ |
| 1120 |
% \begin{rightwordgroup}[curlystyle=\color{blue}]{Sign-extended} |
| 1121 |
% \bitbox{4}{Tag} & \bitbox{12}{Data} |
| 1122 |
% \end{rightwordgroup} \\ |
| 1123 |
% \end{bytefield} |
| 1124 |
% \end{verbatim} |
| 1125 |
% |
| 1126 |
% \begin{bffigure} |
| 1127 |
% \begin{bytefield}{16} |
| 1128 |
% \bitheader{0-15} \\ |
| 1129 |
% \begin{rightwordgroup}[curlystyle=\color{blue}]{Sign-extended} |
| 1130 |
% \bitbox{4}{Tag} & \bitbox{12}{Data} |
| 1131 |
% \end{rightwordgroup} \\ |
| 1132 |
% \end{bytefield} |
| 1133 |
% \end{bffigure} |
| 1134 |
% |
| 1135 |
% \begin{decl} |
| 1136 |
% \optname{lsb} = \meta{integer} |
| 1137 |
% \end{decl} |
| 1138 |
% |
| 1139 |
% Designate the least significant bit (LSB) in the bit header. By |
| 1140 |
% default, the LSB is zero, which means that the first bit position in |
| 1141 |
% the header corresponds to bit~0. Specifying a different LSB shifts |
| 1142 |
% the bit header such that the first bit position instead corresponds to |
| 1143 |
% \meta{integer}. Note that the \optname{lsb} option affects bit |
| 1144 |
% \emph{positions} regardless of whether these positions are labeled, as |
| 1145 |
% demonstrated by the following two examples: |
| 1146 |
% |
| 1147 |
% \begin{verbatim} |
| 1148 |
% \begin{bytefield}{32} |
| 1149 |
% \bitheader[lsb=0]{4,12,20,28} \\ |
| 1150 |
% \bitbox{16}{ar\$hrd} & \bitbox{16}{ar\$pro} \\ |
| 1151 |
% \bitbox{8}{ar\$hln} & \bitbox{8}{ar\$pln} & \bitbox{16}{ar\$op} \\ |
| 1152 |
% \end{bytefield} |
| 1153 |
% \end{verbatim} |
| 1154 |
% |
| 1155 |
% \begin{bffigure} |
| 1156 |
% \begin{bytefield}{32} |
| 1157 |
% \bitheader[lsb=0]{4,12,20,28} \\ |
| 1158 |
% \bitbox{16}{ar\$hrd} & \bitbox{16}{ar\$pro} \\ |
| 1159 |
% \bitbox{8}{ar\$hln} & \bitbox{8}{ar\$pln} & \bitbox{16}{ar\$op} \\ |
| 1160 |
% \end{bytefield} |
| 1161 |
% \end{bffigure} |
| 1162 |
% |
| 1163 |
% \begin{verbatim} |
| 1164 |
% \begin{bytefield}{32} |
| 1165 |
% \bitheader[lsb=4]{4,12,20,28} \\ |
| 1166 |
% \bitbox{16}{ar\$hrd} & \bitbox{16}{ar\$pro} \\ |
| 1167 |
% \bitbox{8}{ar\$hln} & \bitbox{8}{ar\$pln} & \bitbox{16}{ar\$op} \\ |
| 1168 |
% \end{bytefield} |
| 1169 |
% \end{verbatim} |
| 1170 |
% |
| 1171 |
% \begin{bffigure} |
| 1172 |
% \begin{bytefield}{32} |
| 1173 |
% \bitheader[lsb=4]{4,12,20,28} \\ |
| 1174 |
% \bitbox{16}{ar\$hrd} & \bitbox{16}{ar\$pro} \\ |
| 1175 |
% \bitbox{8}{ar\$hln} & \bitbox{8}{ar\$pln} & \bitbox{16}{ar\$op} \\ |
| 1176 |
% \end{bytefield} |
| 1177 |
% \end{bffigure} |
| 1178 |
% |
| 1179 |
% \begin{decl} |
| 1180 |
% \optname{perword} = \meta{command} |
| 1181 |
% \end{decl} |
| 1182 |
% |
| 1183 |
% Provide a macro that will be invoked once for each word in a word box |
| 1184 |
% after the regular content is rendered. The macro will be passed two |
| 1185 |
% arguments: the word number (starting from~0) and the total number of |
| 1186 |
% words in the word box. Furthermore, the macro will be invoked within |
| 1187 |
% a one-word-wide box positioned at the base of the word. |
| 1188 |
% \optname{perword} can therefore be used for delineating words within a |
| 1189 |
% word box, numbering words, or performing other such annotations. As a |
| 1190 |
% simple example, the following code draws a gray line at the bottom of |
| 1191 |
% each word in the ``Descriptive text'' word box: |
| 1192 |
% |
| 1193 |
% \begin{verbatim} |
| 1194 |
% \newcommand{\wordline}[2]{\color[rgb]{0.7,0.7,0.7}\hrulefill} |
| 1195 |
% \begin{bytefield}[bitwidth=4em]{8} |
| 1196 |
% \bitheader[lsb=1,bitformatting=\small]{1-8} \\ |
| 1197 |
% \wordbox[lrt]{7}[perword=\wordline]{Descriptive text (60 bytes)} \\ |
| 1198 |
% \bitbox[lrb]{4}{} & \bitbox{4}{subsys data offset} \\ |
| 1199 |
% \bitbox{4}{subsys data offset} & \bitbox{2}{version} & |
| 1200 |
% \bitbox{2}{endian indicator} \\ |
| 1201 |
% \end{bytefield} |
| 1202 |
% \end{verbatim} |
| 1203 |
% |
| 1204 |
% \begin{bffigure} |
| 1205 |
% \newcommand{\wordline}[2]{\color[rgb]{0.7,0.7,0.7}\hrulefill} |
| 1206 |
% \begin{bytefield}[bitwidth=4em]{8} |
| 1207 |
% \bitheader[lsb=1,bitformatting=\small]{1-8} \\ |
| 1208 |
% \wordbox[lrt]{7}[perword=\wordline]{Descriptive text (60 bytes)} \\ |
| 1209 |
% \bitbox[lrb]{4}{} & \bitbox{4}{subsys data offset} \\ |
| 1210 |
% \bitbox{4}{subsys data offset} & \bitbox{2}{version} & |
| 1211 |
% \bitbox{2}{endian indicator} \\ |
| 1212 |
% \end{bytefield} |
| 1213 |
% \end{bffigure} |
| 1214 |
% |
| 1215 |
% |
| 1216 |
% \subsection{Common tricks} |
| 1217 |
% \label{sec:tricks} |
| 1218 |
% |
| 1219 |
% This section shows some clever ways to use \pkgname{bytefield}'s commands |
| 1220 |
% to produce some useful effects. |
| 1221 |
% |
| 1222 |
% \paragraph{Odd-sized fields} |
| 1223 |
% To produce a field that is, say, 1\textonehalf{} words long, use a |
| 1224 |
% |\bitbox| for the fractional part and specify appropriate values |
| 1225 |
% for the various \meta{sides} parameters. For instance: |
| 1226 |
% |
| 1227 |
% \begin{verbatim} |
| 1228 |
% \begin{bytefield}{16} |
| 1229 |
% \bitheader{0,7,8,15} \\ |
| 1230 |
% \bitbox{8}{8-bit field} & \bitbox[lrt]{8}{} \\ |
| 1231 |
% \wordbox[lrb]{1}{24-bit field} |
| 1232 |
% \end{bytefield} |
| 1233 |
% \end{verbatim} |
| 1234 |
% |
| 1235 |
% \begin{bffigure} |
| 1236 |
% \begin{bytefield}{16} |
| 1237 |
% \bitheader{0,7,8,15} \\ |
| 1238 |
% \bitbox{8}{8-bit field} & \bitbox[lrt]{8}{} \\ |
| 1239 |
% \wordbox[lrb]{1}{24-bit field} |
| 1240 |
% \end{bytefield} |
| 1241 |
% \end{bffigure} |
| 1242 |
% |
| 1243 |
% \paragraph{Ellipses} |
| 1244 |
% To skip words that appear the middle of enumerated data, put some |
| 1245 |
% |\vdots| in a |\wordbox| with empty \meta{sides}: |
| 1246 |
% |
| 1247 |
% \begin{verbatim} |
| 1248 |
% \begin{bytefield}{16} |
| 1249 |
% \bitbox{8}{Type} & \bitbox{8}{\# of nodes} \\ |
| 1250 |
% \wordbox{1}{Node~1} \\ |
| 1251 |
% \wordbox{1}{Node~2} \\ |
| 1252 |
% \wordbox[]{1}{$\vdots$} \\[1ex] |
| 1253 |
% \wordbox{1}{Node~$N$} |
| 1254 |
% \end{bytefield} |
| 1255 |
% \end{verbatim} |
| 1256 |
% |
| 1257 |
% \begin{bffigure} |
| 1258 |
% \begin{bytefield}{16} |
| 1259 |
% \bitbox{8}{Type} & \bitbox{8}{\# of nodes} \\ |
| 1260 |
% \wordbox{1}{Node~1} \\ |
| 1261 |
% \wordbox{1}{Node~2} \\ |
| 1262 |
% \wordbox[]{1}{$\vdots$} \\[1ex] |
| 1263 |
% \wordbox{1}{Node~$N$} |
| 1264 |
% \end{bytefield} |
| 1265 |
% \end{bffigure} |
| 1266 |
% |
| 1267 |
% \noindent |
| 1268 |
% The extra |1ex| of vertical space helps vertically center the |\vdots| |
| 1269 |
% a bit better. |
| 1270 |
% |
| 1271 |
% \paragraph{Narrow fields} |
| 1272 |
% There are a number of options for labeling a narrow field (e.g.,~one |
| 1273 |
% occupying a single bit): |
| 1274 |
% |
| 1275 |
% \newsavebox{\defaultOK} |
| 1276 |
% \begin{lrbox}{\defaultOK} |
| 1277 |
% \begin{bytefield}{8} |
| 1278 |
% \bitbox{1}{OK} & \bitbox{7}{Data} \\ |
| 1279 |
% \end{bytefield} |
| 1280 |
% \end{lrbox} |
| 1281 |
% |
| 1282 |
% \newsavebox{\bitwidthOK} |
| 1283 |
% \begin{lrbox}{\bitwidthOK} |
| 1284 |
% \bytefieldsetup{bitwidth=\widthof{OK~}}\relax |
| 1285 |
% \begin{bytefield}{8} |
| 1286 |
% \bitbox{1}{OK} & \bitbox{7}{Data} \\ |
| 1287 |
% \end{bytefield} |
| 1288 |
% \end{lrbox} |
| 1289 |
% |
| 1290 |
% \newsavebox{\tinyOK} |
| 1291 |
% \begin{lrbox}{\tinyOK} |
| 1292 |
% \begin{bytefield}{8} |
| 1293 |
% \bitbox{1}{\tiny OK} & \bitbox{7}{Data} \\ |
| 1294 |
% \end{bytefield} |
| 1295 |
% \end{lrbox} |
| 1296 |
% |
| 1297 |
% \newsavebox{\verticalOK} |
| 1298 |
% \begin{lrbox}{\verticalOK} |
| 1299 |
% \begin{bytefield}{8} |
| 1300 |
% \bitbox{1}{\tiny O \\ K} & \bitbox{7}{Data} \\ |
| 1301 |
% \end{bytefield} |
| 1302 |
% \end{lrbox} |
| 1303 |
% |
| 1304 |
% \newsavebox{\rotateOK} |
| 1305 |
% \begin{lrbox}{\rotateOK} |
| 1306 |
% \begin{bytefield}{8} |
| 1307 |
% \bitbox{1}{\rotatebox{90}{\small OK}} & \bitbox{7}{Data} \\ |
| 1308 |
% \end{bytefield} |
| 1309 |
% \end{lrbox} |
| 1310 |
% |
| 1311 |
% \newsavebox{\resizeOK} |
| 1312 |
% \begin{lrbox}{\resizeOK} |
| 1313 |
% \begin{bytefield}{8} |
| 1314 |
% \bitbox{1}{\let\bw=\width\resizebox{\bw}{!}{~OK~}} & \bitbox{7}{Data} \\ |
| 1315 |
% \end{bytefield} |
| 1316 |
% \end{lrbox} |
| 1317 |
% |
| 1318 |
% \bigskip |
| 1319 |
% |
| 1320 |
% \begin{tabular}{ll} |
| 1321 |
% \emph{Default}: & \raisebox{-4ex}{\usebox{\defaultOK}} \\ |
| 1322 |
% |\bytefieldsetup{%| |
| 1323 |
% & \multirow{2}*{\raisebox{-4ex}{\usebox{\bitwidthOK}}} \\ |
| 1324 |
% | bitwidth=\widthof{OK~}}|: & \\[2ex] |
| 1325 |
% |\tiny OK|: & \raisebox{-4ex}{\usebox{\tinyOK}} \\ |
| 1326 |
% |\tiny O \\ K|: & \raisebox{-4ex}{\usebox{\verticalOK}} \\ |
| 1327 |
% |\rotatebox{90}{\small OK}|: & \raisebox{-4ex}{\usebox{\rotateOK}} \\ |
| 1328 |
% |\let\bw=\width| & \multirow{2}*{\raisebox{-4ex}{\usebox{\resizeOK}}} \\ |
| 1329 |
% |\resizebox{\bw}{!}{~OK~}|: & \\[2ex] |
| 1330 |
% \end{tabular} |
| 1331 |
% |
| 1332 |
% \paragraph{Multi-line bit fields} |
| 1333 |
% Presentations of wide registers are often easier to read when split |
| 1334 |
% across multiple lines. (This capability was originally requested by |
| 1335 |
% Chris L'Esperance and is currently implemented in \pkgname{bytefield} |
| 1336 |
% based on code provided by Renaud Pacalet.) The trick behind the |
| 1337 |
% typesetting of multi-line bit fields is to pass the \optname{lsb} |
| 1338 |
% option to |\bitheader| to change the starting bit number used in each |
| 1339 |
% bit header: |
| 1340 |
% |
| 1341 |
% \begin{verbatim} |
| 1342 |
% \begin{bytefield}[endianness=big,bitwidth=2em]{16} |
| 1343 |
% \bitheader[lsb=16]{16-31} \\ |
| 1344 |
% \bitbox{1}{\tiny Enable} & \bitbox{7}{Reserved} |
| 1345 |
% & \bitbox{8}{Bus} \\[3ex] |
| 1346 |
% \bitheader{0-15} \\ |
| 1347 |
% \bitbox{5}{Device} & \bitbox{3}{Function} & \bitbox{6}{Register} |
| 1348 |
% & \bitbox{2}{00} |
| 1349 |
% \end{bytefield} |
| 1350 |
% \end{verbatim} |
| 1351 |
% |
| 1352 |
% \begin{bffigure} |
| 1353 |
% \begin{bytefield}[endianness=big,bitwidth=2em]{16} |
| 1354 |
% \bitheader[lsb=16]{16-31} \\ |
| 1355 |
% \bitbox{1}{\tiny Enable} & \bitbox{7}{Reserved} |
| 1356 |
% & \bitbox{8}{Bus} \\[3ex] |
| 1357 |
% \bitheader{0-15} \\ |
| 1358 |
% \bitbox{5}{Device} & \bitbox{3}{Function} & \bitbox{6}{Register} |
| 1359 |
% & \bitbox{2}{00} |
| 1360 |
% \end{bytefield} |
| 1361 |
% \end{bffigure} |
| 1362 |
% |
| 1363 |
% Note the use of the optional argument to |\\| to introduce three |
| 1364 |
% \mbox{x-heights} of additional whitespace between the two rows of |
| 1365 |
% bits. |
| 1366 |
% |
| 1367 |
% \paragraph{Rotated bit labels} |
| 1368 |
% A problem with using very large bit numbers is that the labels run |
| 1369 |
% into each other, as in the following example: |
| 1370 |
% |
| 1371 |
% \begin{verbatim} |
| 1372 |
% \begin{bytefield}[endianness=big]{8} |
| 1373 |
% \bitheader[lsb=995]{995-1002} \\ |
| 1374 |
% \bitbox{4}{A} & \bitbox{4}{B} |
| 1375 |
% \end{bytefield} |
| 1376 |
% \end{verbatim} |
| 1377 |
% |
| 1378 |
% \begin{bffigure} |
| 1379 |
% \begin{bytefield}[endianness=big]{8} |
| 1380 |
% \bitheader[lsb=995]{995-1002} \\ |
| 1381 |
% \bitbox{4}{A} & \bitbox{4}{B} |
| 1382 |
% \end{bytefield} |
| 1383 |
% \end{bffigure} |
| 1384 |
% |
| 1385 |
% One solution is to use the \optname{bitformatting} option and the |
| 1386 |
% \pkgname{graphicx} package's |\rotatebox| command to rotate each bit |
| 1387 |
% label by 90\textdegree. Unfortunately, the naive use of |
| 1388 |
% \optname{bitformatting} and |\rotatebox| does not typeset nicely: |
| 1389 |
% |
| 1390 |
% \begin{verbatim} |
| 1391 |
% \begin{bytefield}[endianness=big]{8} |
| 1392 |
% \bitheader[lsb=995, |
| 1393 |
% bitformatting={\tiny\rotatebox[origin=B]{90}}]{995-1002} \\ |
| 1394 |
% \bitbox{4}{A} & \bitbox{4}{B} |
| 1395 |
% \end{bytefield} |
| 1396 |
% \end{verbatim} |
| 1397 |
% |
| 1398 |
% \begin{bffigure} |
| 1399 |
% \begin{bytefield}[endianness=big]{8} |
| 1400 |
% \bitheader[lsb=995, |
| 1401 |
% bitformatting={\tiny\rotatebox[origin=B]{90}}]{995-1002} \\ |
| 1402 |
% \bitbox{4}{A} & \bitbox{4}{B} |
| 1403 |
% \end{bytefield} |
| 1404 |
% \end{bffigure} |
| 1405 |
% |
| 1406 |
% The two problems are that (1)~the numbers are left-justified, and |
| 1407 |
% (2)~the numbers touch the top margin of the word box. To address these |
| 1408 |
% problems we use |\makebox| to construct a right-justified region that |
| 1409 |
% is sufficiently wide to hold our largest number plus some additional |
| 1410 |
% space to shift the rotated numbers upwards: |
| 1411 |
% |
| 1412 |
% \begin{verbatim} |
| 1413 |
% \newlength{\bitlabelwidth} |
| 1414 |
% \newcommand{\rotbitheader}[1]{% |
| 1415 |
% \tiny |
| 1416 |
% \settowidth{\bitlabelwidth}{\quad 9999}% |
| 1417 |
% \rotatebox[origin=B]{90}{\makebox[\bitlabelwidth][r]{#1}}% |
| 1418 |
% } |
| 1419 |
% |
| 1420 |
% \begin{bytefield}[endianness=big]{8} |
| 1421 |
% \bitheader[lsb=995,bitformatting=\rotbitheader]{995-1002} \\ |
| 1422 |
% \bitbox{4}{A} & \bitbox{4}{B} |
| 1423 |
% \end{bytefield} |
| 1424 |
% \end{verbatim} |
| 1425 |
% |
| 1426 |
% \begin{bffigure} |
| 1427 |
% \newlength{\bitlabelwidth} |
| 1428 |
% \newcommand{\rotbitheader}[1]{^^A |
| 1429 |
% \tiny |
| 1430 |
% \settowidth{\bitlabelwidth}{\quad 9999}% |
| 1431 |
% \rotatebox[origin=B]{90}{\makebox[\bitlabelwidth][r]{#1}}^^A |
| 1432 |
% } |
| 1433 |
% |
| 1434 |
% \begin{bytefield}[endianness=big]{8} |
| 1435 |
% \bitheader[lsb=995,bitformatting=\rotbitheader]{995-1002} \\ |
| 1436 |
% \bitbox{4}{A} & \bitbox{4}{B} |
| 1437 |
% \end{bytefield} |
| 1438 |
% \end{bffigure} |
| 1439 |
% |
| 1440 |
% \paragraph{Unused bits} |
| 1441 |
% The \optname{bgcolor} option can be used to represent unused bits by |
| 1442 |
% specifying a background fill color---light gray looks nice---and empty |
| 1443 |
% text: |
| 1444 |
% |
| 1445 |
% \begin{verbatim} |
| 1446 |
% \definecolor{lightgray}{gray}{0.8} |
| 1447 |
% \begin{bytefield}{32} |
| 1448 |
% \bitheader{0,4,8,12,16,20,24,28} \\ |
| 1449 |
% \bitbox{8}{Tag} & \bitbox{8}{Value} & |
| 1450 |
% \bitbox{4}[bgcolor=lightgray]{} & |
| 1451 |
% \bitbox{12}{Mask} \\ |
| 1452 |
% \wordbox{1}{Key} |
| 1453 |
% \end{bytefield} |
| 1454 |
% \end{verbatim} |
| 1455 |
% |
| 1456 |
% \definecolor{lightgray}{gray}{0.8} |
| 1457 |
% \begin{bffigure} |
| 1458 |
% \begin{bytefield}{32} |
| 1459 |
% \bitheader{0,4,8,12,16,20,24,28} \\ |
| 1460 |
% \bitbox{8}{Tag} & \bitbox{8}{Value} & |
| 1461 |
% \bitbox{4}[bgcolor=lightgray]{} & |
| 1462 |
% \bitbox{12}{Mask} \\ |
| 1463 |
% \wordbox{1}{Key} |
| 1464 |
% \end{bytefield} |
| 1465 |
% \end{bffigure} |
| 1466 |
% |
| 1467 |
% \paragraph{Aligning text on the baseline} |
| 1468 |
% Because \pkgname{bytefield} internally uses \LaTeX's |picture| |
| 1469 |
% environment and that environment's |\makebox| command to draw bit |
| 1470 |
% boxes and word boxes, the text within a box is centered vertically |
| 1471 |
% with no attention paid to the text's baseline. As a result, some |
| 1472 |
% bit-field labels appear somewhat askew: |
| 1473 |
% |
| 1474 |
% \begin{verbatim} |
| 1475 |
% \begin{bytefield}[bitwidth=1.5em]{2} |
| 1476 |
% \bitbox{1}{M} & \bitbox{1}{y} |
| 1477 |
% \end{bytefield} |
| 1478 |
% \end{verbatim} |
| 1479 |
% |
| 1480 |
% \begin{bffigure} |
| 1481 |
% \begin{bytefield}[bitwidth=1.5em]{2} |
| 1482 |
% \bitbox{1}{M} & \bitbox{1}{y} |
| 1483 |
% \end{bytefield} |
| 1484 |
% \end{bffigure} |
| 1485 |
% |
| 1486 |
% A solution is to use the \optname{boxformatting} option to trick |
| 1487 |
% |\makebox| into thinking that all text has the same height and depth. |
| 1488 |
% Here we use |\raisebox| to indicate that all text is as tall as a |
| 1489 |
% ``|W|'' and does not descend at all below the baseline: |
| 1490 |
% |
| 1491 |
% \begin{verbatim} |
| 1492 |
% \newlength{\maxheight} |
| 1493 |
% \setlength{\maxheight}{\heightof{W}} |
| 1494 |
% |
| 1495 |
% \newcommand{\baselinealign}[1]{% |
| 1496 |
% \centering |
| 1497 |
% \raisebox{0pt}[\maxheight][0pt]{#1}% |
| 1498 |
% } |
| 1499 |
% |
| 1500 |
% \begin{bytefield}[boxformatting=\baselinealign, |
| 1501 |
% bitwidth=1.5em]{2} |
| 1502 |
% \bitbox{1}{M} & \bitbox{1}{y} |
| 1503 |
% \end{bytefield} |
| 1504 |
% \end{verbatim} |
| 1505 |
% |
| 1506 |
% \begin{bffigure} |
| 1507 |
% \newlength{\maxheight} |
| 1508 |
% \setlength{\maxheight}{\heightof{W}} |
| 1509 |
% \newcommand{\baselinealign}[1]{^^A |
| 1510 |
% \centering |
| 1511 |
% \raisebox{0pt}[\maxheight][0pt]{#1}^^A |
| 1512 |
% } |
| 1513 |
% \begin{bytefield}[boxformatting=\baselinealign, |
| 1514 |
% bitwidth=1.5em]{2} |
| 1515 |
% \bitbox{1}{M} & \bitbox{1}{y} |
| 1516 |
% \end{bytefield} |
| 1517 |
% \end{bffigure} |
| 1518 |
% |
| 1519 |
% \paragraph{Register contents} |
| 1520 |
% Sometimes, rather than listing the \emph{meaning} of each bit field |
| 1521 |
% within each |\bitbox| or |\wordbox|, it may be desirable to list the |
| 1522 |
% \emph{contents}, with the meaning described in an additional label |
| 1523 |
% above each bit number in the bit header. Although the |
| 1524 |
% \pkgname{register} package is more suited to this form of layout, |
| 1525 |
% \pkgname{bytefield} can serve in a pinch with the help of the |
| 1526 |
% |\turnbox| macro from the \pkgname{rotating} package: |
| 1527 |
% |
| 1528 |
% \begin{verbatim} |
| 1529 |
% \newcommand{\bitlabel}[2]{% |
| 1530 |
% \bitbox[]{#1}{% |
| 1531 |
% \raisebox{0pt}[4ex][0pt]{% |
| 1532 |
% \turnbox{45}{\fontsize{7}{7}\selectfont#2}% |
| 1533 |
% }% |
| 1534 |
% }% |
| 1535 |
% } |
| 1536 |
% |
| 1537 |
% \begin{bytefield}[bitwidth=1em]{16} |
| 1538 |
% \bitlabel{1}{Carry} & \bitlabel{1}{Reserved} & |
| 1539 |
% \bitlabel{1}{Parity} & \bitlabel{1}{Reserved} & |
| 1540 |
% \bitlabel{1}{Adjust} & \bitlabel{1}{Reserved} & |
| 1541 |
% \bitlabel{1}{Zero} & \bitlabel{1}{Sign} & |
| 1542 |
% \bitlabel{1}{Trap} & \bitlabel{1}{Interrupt enable} & |
| 1543 |
% \bitlabel{1}{Direction} & \bitlabel{1}{Overflow} & |
| 1544 |
% \bitlabel{2}{I/O privilege level (12--13)} & |
| 1545 |
% \bitlabel{1}{Nested task} & \bitlabel{1}{Reserved} \\ |
| 1546 |
% \bitheader{0-15} \\ |
| 1547 |
% \bitbox{1}{0} & \bitbox{1}{1} & \bitbox{1}{0} & \bitbox{1}{0} & |
| 1548 |
% \bitbox{1}{0} & \bitbox{1}{0} & \bitbox{1}{0} & \bitbox{1}{1} & |
| 1549 |
% \bitbox{1}{0} & \bitbox{1}{1} & \bitbox{1}{0} & \bitbox{1}{0} & |
| 1550 |
% \bitbox{1}{0} & \bitbox{1}{0} & \bitbox{1}{0} & \bitbox{1}{0} |
| 1551 |
% \end{bytefield} |
| 1552 |
% \end{verbatim} |
| 1553 |
% |
| 1554 |
% \vspace{3\baselineskip} ^^A Allow extra room for the rotated headers. |
| 1555 |
% \begin{bffigure} |
| 1556 |
% \newcommand{\bitlabel}[2]{% |
| 1557 |
% \bitbox[]{#1}{% |
| 1558 |
% \raisebox{0pt}[4ex][0pt]{% |
| 1559 |
% \turnbox{45}{\fontsize{7}{7}\selectfont#2}% |
| 1560 |
% }% |
| 1561 |
% }% |
| 1562 |
% } |
| 1563 |
% \begin{bytefield}[bitwidth=1em]{16} |
| 1564 |
% \bitlabel{1}{Carry} & \bitlabel{1}{Reserved} & |
| 1565 |
% \bitlabel{1}{Parity} & \bitlabel{1}{Reserved} & |
| 1566 |
% \bitlabel{1}{Adjust} & \bitlabel{1}{Reserved} & |
| 1567 |
% \bitlabel{1}{Zero} & \bitlabel{1}{Sign} & |
| 1568 |
% \bitlabel{1}{Trap} & \bitlabel{1}{Interrupt enable} & |
| 1569 |
% \bitlabel{1}{Direction} & \bitlabel{1}{Overflow} & |
| 1570 |
% \bitlabel{2}{I/O privilege level (12--13)} & |
| 1571 |
% \bitlabel{1}{Nested task} & \bitlabel{1}{Reserved} \\ |
| 1572 |
% \bitheader{0-15} \\ |
| 1573 |
% \bitbox{1}{0} & \bitbox{1}{1} & \bitbox{1}{0} & \bitbox{1}{0} & |
| 1574 |
% \bitbox{1}{0} & \bitbox{1}{0} & \bitbox{1}{0} & \bitbox{1}{1} & |
| 1575 |
% \bitbox{1}{0} & \bitbox{1}{1} & \bitbox{1}{0} & \bitbox{1}{0} & |
| 1576 |
% \bitbox{1}{0} & \bitbox{1}{0} & \bitbox{1}{0} & \bitbox{1}{0} |
| 1577 |
% \end{bytefield} |
| 1578 |
% \end{bffigure} |
| 1579 |
% |
| 1580 |
% |
| 1581 |
% \subsection{Not-so-common tricks} |
| 1582 |
% |
| 1583 |
% \paragraph{Omitted bit numbers} |
| 1584 |
% It is occasionally convenient to show a wide bit field in which the |
| 1585 |
% middle numbers are replaced with an ellipsis. The trick to |
| 1586 |
% typesetting such a thing with \pkgname{bytefield} is to point the |
| 1587 |
% |bitformatting| option to a macro that conditionally modifies the |
| 1588 |
% given bit number before outputting it. One catch is that |
| 1589 |
% \pkgname{bytefield} measures the height of the string ``|1234567890|'' |
| 1590 |
% using the current bit formatting, so that needs to be a valid input. |
| 1591 |
% (If \optname{bitwidth} is set to ``|auto|'', then ``|99i|'' also has |
| 1592 |
% to be a valid input, but we're not using ``|auto|'' here.) The |
| 1593 |
% following example shows how to \emph{conditionally} modify the bit |
| 1594 |
% number: If the number is |1234567890|, it is used as is; numbers |
| 1595 |
% greater than~9 are increased by 48; numbers less than~4 are |
| 1596 |
% unmodified; the number~6 is replaced by an ellipsis; and all other |
| 1597 |
% numbers are discarded. |
| 1598 |
% |
| 1599 |
% \begin{verbatim} |
| 1600 |
% \newcommand{\fakesixtyfourbits}[1]{% |
| 1601 |
% \tiny |
| 1602 |
% \ifnum#1=1234567890 |
| 1603 |
% #1 |
| 1604 |
% \else |
| 1605 |
% \ifnum#1>9 |
| 1606 |
% \count32=#1 |
| 1607 |
% \advance\count32 by 48 |
| 1608 |
% \the\count32% |
| 1609 |
% \else |
| 1610 |
% \ifnum#1<4 |
| 1611 |
% #1% |
| 1612 |
% \else |
| 1613 |
% \ifnum#1=6 |
| 1614 |
% $\cdots$% |
| 1615 |
% \fi |
| 1616 |
% \fi |
| 1617 |
% \fi |
| 1618 |
% \fi |
| 1619 |
% } |
| 1620 |
% \begin{bytefield}[% |
| 1621 |
% bitwidth=\widthof{\tiny Fwd~}, |
| 1622 |
% bitformatting=\fakesixtyfourbits, |
| 1623 |
% endianness=big]{16} |
| 1624 |
% \bitheader{0-15} \\ |
| 1625 |
% \bitbox{1}{\tiny F/E} & \bitbox{1}{\tiny T0} & \bitbox{1}{\tiny T1} |
| 1626 |
% & \bitbox{1}{\tiny Fwd} & \bitbox{12}{Data value} |
| 1627 |
% \end{bytefield} |
| 1628 |
% \end{verbatim} |
| 1629 |
% |
| 1630 |
% \begin{bffigure} |
| 1631 |
% \newcommand{\fakesixtyfourbits}[1]{% |
| 1632 |
% \tiny |
| 1633 |
% \ifnum#1=1234567890 |
| 1634 |
% #1 |
| 1635 |
% \else |
| 1636 |
% \ifnum#1>9 |
| 1637 |
% \count32=#1 |
| 1638 |
% \advance\count32 by 48 |
| 1639 |
% \the\count32^^A |
| 1640 |
% \else |
| 1641 |
% \ifnum#1<4 |
| 1642 |
% #1% |
| 1643 |
% \else |
| 1644 |
% \ifnum#1=6 |
| 1645 |
% $\cdots$^^A |
| 1646 |
% \fi |
| 1647 |
% \fi |
| 1648 |
% \fi |
| 1649 |
% \fi |
| 1650 |
% } |
| 1651 |
% \begin{bytefield}[^^A |
| 1652 |
% bitwidth=\widthof{\tiny Fwd~}, |
| 1653 |
% bitformatting=\fakesixtyfourbits, |
| 1654 |
% endianness=big]{16} |
| 1655 |
% \bitheader{0-15} \\ |
| 1656 |
% \bitbox{1}{\tiny F/E} & \bitbox{1}{\tiny T0} & \bitbox{1}{\tiny T1} |
| 1657 |
% & \bitbox{1}{\tiny Fwd} & \bitbox{12}{Data value} |
| 1658 |
% \end{bytefield} |
| 1659 |
% \end{bffigure} |
| 1660 |
% |
| 1661 |
% \paragraph{Memory-map diagrams} |
| 1662 |
% While certainly not the intended purpose of the \pkgname{bytefield} |
| 1663 |
% package, one can utilize word boxes with empty \meta{sides} and |
| 1664 |
% word labels to produce memory-map diagrams: |
| 1665 |
% |
| 1666 |
% \begin{verbatim} |
| 1667 |
% \newcommand{\descbox}[2]{\parbox[c][3.8\baselineskip]{0.95\width}{% |
| 1668 |
% \raggedright #1\vfill #2}} |
| 1669 |
% \begin{bytefield}[bitheight=4\baselineskip]{32} |
| 1670 |
% \begin{rightwordgroup}{Partition 4} |
| 1671 |
% \bitbox[]{8}{\texttt{0xFFFFFFFF} \\[2\baselineskip] |
| 1672 |
% \texttt{0xC0000000}} & |
| 1673 |
% \bitbox{24}{\descbox{1\,GB area for VxDs, memory manager, |
| 1674 |
% file system code; shared by all processes.}{Read/writable.}} |
| 1675 |
% \end{rightwordgroup} \\ |
| 1676 |
% \begin{rightwordgroup}{Partition 3} |
| 1677 |
% \bitbox[]{8}{\texttt{0xBFFFFFFF} \\[2\baselineskip] |
| 1678 |
% \texttt{0x80000000}} & |
| 1679 |
% \bitbox{24}{\descbox{1\,GB area for memory-mapped files, |
| 1680 |
% shared system \textsc{dll}s, file system code; shared by all |
| 1681 |
% processes.}{Read/writable.}} |
| 1682 |
% \end{rightwordgroup} \\ |
| 1683 |
% \begin{rightwordgroup}{Partition 2} |
| 1684 |
% \bitbox[]{8}{\texttt{0x7FFFFFFF} \\[2\baselineskip] |
| 1685 |
% \texttt{0x00400000}} & |
| 1686 |
% \bitbox{24}{\descbox{$\sim$2\,GB area private to process, |
| 1687 |
% process code, and data.}{Read/writable.}} |
| 1688 |
% \end{rightwordgroup} \\ |
| 1689 |
% \begin{rightwordgroup}{Partition 1} |
| 1690 |
% \bitbox[]{8}{\texttt{0x003FFFFF} \\[2\baselineskip] |
| 1691 |
% \texttt{0x00001000}} & |
| 1692 |
% \bitbox{24}{\descbox{4\,MB area for MS-DOS and Windows~3.1 |
| 1693 |
% compatibility.}{Read/writable.}} \\ |
| 1694 |
% \bitbox[]{8}{\texttt{0x00000FFF} \\[2\baselineskip] |
| 1695 |
% \texttt{0x00000000}} & |
| 1696 |
% \bitbox{24}{\descbox{4096~byte area for MS-DOS and Windows~3.1 |
| 1697 |
% compatibility.}{Protected---catches \textsc{null} |
| 1698 |
% pointers.}} |
| 1699 |
% \end{rightwordgroup} |
| 1700 |
% \end{bytefield} |
| 1701 |
% \end{verbatim} |
| 1702 |
% |
| 1703 |
% \begin{bffigure} |
| 1704 |
% \newcommand{\descbox}[2]{\parbox[c][3.8\baselineskip]{0.95\width}{% |
| 1705 |
% \raggedright #1\vfill #2}} |
| 1706 |
% \begin{bytefield}[bitheight=4\baselineskip]{32} |
| 1707 |
% \begin{rightwordgroup}{Partition 4} |
| 1708 |
% \bitbox[]{8}{\texttt{0xFFFFFFFF} \\[2\baselineskip] |
| 1709 |
% \texttt{0xC0000000}} & |
| 1710 |
% \bitbox{24}{\descbox{1\,GB area for VxDs, memory manager, |
| 1711 |
% file system code; shared by all processes.}{Read/writable.}} |
| 1712 |
% \end{rightwordgroup} \\ |
| 1713 |
% \begin{rightwordgroup}{Partition 3} |
| 1714 |
% \bitbox[]{8}{\texttt{0xBFFFFFFF} \\[2\baselineskip] |
| 1715 |
% \texttt{0x80000000}} & |
| 1716 |
% \bitbox{24}{\descbox{1\,GB area for memory-mapped files, |
| 1717 |
% shared system \textsc{dll}s, file system code; shared by all |
| 1718 |
% processes.}{Read/writable.}} |
| 1719 |
% \end{rightwordgroup} \\ |
| 1720 |
% \begin{rightwordgroup}{Partition 2} |
| 1721 |
% \bitbox[]{8}{\texttt{0x7FFFFFFF} \\[2\baselineskip] |
| 1722 |
% \texttt{0x00400000}} & |
| 1723 |
% \bitbox{24}{\descbox{$\sim$2\,GB area private to process, process |
| 1724 |
% code, and data.}{Read/writable.}} |
| 1725 |
% \end{rightwordgroup} \\ |
| 1726 |
% \begin{rightwordgroup}{Partition 1} |
| 1727 |
% \bitbox[]{8}{\texttt{0x003FFFFF} \\[2\baselineskip] |
| 1728 |
% \texttt{0x00001000}} & |
| 1729 |
% \bitbox{24}{\descbox{4\,MB area for MS-DOS and Windows~3.1 |
| 1730 |
% compatibility.}{Read/writable.}} \\ |
| 1731 |
% \bitbox[]{8}{\texttt{0x00000FFF} \\[2\baselineskip] |
| 1732 |
% \texttt{0x00000000}} & |
| 1733 |
% \bitbox{24}{\descbox{4096~byte area for MS-DOS and Windows~3.1 |
| 1734 |
% compatibility.}{Protected---catches \textsc{null} |
| 1735 |
% pointers.}} |
| 1736 |
% \end{rightwordgroup} |
| 1737 |
% \end{bytefield} |
| 1738 |
% \end{bffigure} |
| 1739 |
% |
| 1740 |
% The following variation uses variable-height regions in the memory map: |
| 1741 |
% \changes{v2.1}{2012/09/02}{Included in the documentation a variable-height |
| 1742 |
% memory-map example suggested by Martin Demling} |
| 1743 |
% |
| 1744 |
% \begin{verbatim} |
| 1745 |
% \newcommand{\memsection}[4]{% |
| 1746 |
% % define the height of the memsection |
| 1747 |
% \bytefieldsetup{bitheight=#3\baselineskip}% |
| 1748 |
% \bitbox[]{10}{% |
| 1749 |
% \texttt{#1}% print end address |
| 1750 |
% \\ |
| 1751 |
% % do some spacing |
| 1752 |
% \vspace{#3\baselineskip} |
| 1753 |
% \vspace{-2\baselineskip} |
| 1754 |
% \vspace{-#3pt} |
| 1755 |
% \texttt{#2}% print start address |
| 1756 |
% }% |
| 1757 |
% \bitbox{16}{#4}% print box with caption |
| 1758 |
% } |
| 1759 |
% |
| 1760 |
% \begin{bytefield}{24} |
| 1761 |
% \memsection{ffff ffff}{0040 0000}{15}{-- free --}\\ |
| 1762 |
% \begin{rightwordgroup}{internal memory} |
| 1763 |
% \memsection{003f ffff}{002f c000}{4}{Special Function |
| 1764 |
% Registers}\\ |
| 1765 |
% \memsection{002f bfff}{0007 0000}{3}{-- reserved --}\\ |
| 1766 |
% \memsection{0006 ffff}{0000 0000}{8}{Internal Flash} |
| 1767 |
% \end{rightwordgroup}\\ |
| 1768 |
% \end{bytefield} |
| 1769 |
% \end{verbatim} |
| 1770 |
% |
| 1771 |
% \begin{bffigure} |
| 1772 |
% \newcommand{\memsection}[4]{^^A |
| 1773 |
% \bytefieldsetup{bitheight=#3\baselineskip}^^A |
| 1774 |
% \bitbox[]{10}{^^A |
| 1775 |
% \texttt{#1}^^A |
| 1776 |
% \\ \vspace{#3\baselineskip}\vspace{-2\baselineskip}\vspace{-#3pt}^^A |
| 1777 |
% \texttt{#2}^^A |
| 1778 |
% }^^A |
| 1779 |
% \bitbox{16}{#4}^^A |
| 1780 |
% } |
| 1781 |
% \begin{bytefield}{24} |
| 1782 |
% \memsection{ffff ffff}{0040 0000}{15}{-- free --}\\ |
| 1783 |
% \begin{rightwordgroup}{internal memory} |
| 1784 |
% \memsection{003f ffff}{002f c000}{4}{Special Function Registers}\\ |
| 1785 |
% \memsection{002f bfff}{0007 0000}{3}{-- reserved --}\\ |
| 1786 |
% \memsection{0006 ffff}{0000 0000}{8}{Internal Flash} |
| 1787 |
% \end{rightwordgroup}\\ |
| 1788 |
% \end{bytefield} |
| 1789 |
% \end{bffigure} |
| 1790 |
% |
| 1791 |
% |
| 1792 |
% \subsection{Putting it all together} |
| 1793 |
% |
| 1794 |
% The following code showcases most of \pkgname{bytefield}'s features in a |
| 1795 |
% single figure. |
| 1796 |
% |
| 1797 |
% \begin{verbatim} |
| 1798 |
% \begin{bytefield}[bitheight=2.5\baselineskip]{32} |
| 1799 |
% \bitheader{0,7,8,15,16,23,24,31} \\ |
| 1800 |
% \begin{rightwordgroup}{\parbox{6em}{\raggedright These words were taken |
| 1801 |
% verbatim from the TCP header definition (RFC~793).}} |
| 1802 |
% \bitbox{4}{Data offset} & \bitbox{6}{Reserved} & |
| 1803 |
% \bitbox{1}{\tiny U\\R\\G} & \bitbox{1}{\tiny A\\C\\K} & |
| 1804 |
% \bitbox{1}{\tiny P\\S\\H} & \bitbox{1}{\tiny R\\S\\T} & |
| 1805 |
% \bitbox{1}{\tiny S\\Y\\N} & \bitbox{1}{\tiny F\\I\\N} & |
| 1806 |
% \bitbox{16}{Window} \\ |
| 1807 |
% \bitbox{16}{Checksum} & \bitbox{16}{Urgent pointer} |
| 1808 |
% \end{rightwordgroup} \\ |
| 1809 |
% \wordbox[lrt]{1}{Data octets} \\ |
| 1810 |
% \skippedwords \\ |
| 1811 |
% \wordbox[lrb]{1}{} \\ |
| 1812 |
% \begin{leftwordgroup}{\parbox{6em}{\raggedright Note that we can display, |
| 1813 |
% for example, a misaligned 64-bit value with clever use of the |
| 1814 |
% optional argument to \texttt{\string\wordbox} and |
| 1815 |
% \texttt{\string\bitbox}.}} |
| 1816 |
% \bitbox{8}{Source} & \bitbox{8}{Destination} & |
| 1817 |
% \bitbox[lrt]{16}{} \\ |
| 1818 |
% \wordbox[lr]{1}{Timestamp} \\ |
| 1819 |
% \begin{rightwordgroup}{\parbox{6em}{\raggedright Why two Length fields? |
| 1820 |
% No particular reason.}} |
| 1821 |
% \bitbox[lrb]{16}{} & \bitbox{16}{Length} |
| 1822 |
% \end{leftwordgroup} \\ |
| 1823 |
% \bitbox{6}{Key} & \bitbox{6}{Value} & \bitbox{4}{Unused} & |
| 1824 |
% \bitbox{16}{Length} |
| 1825 |
% \end{rightwordgroup} \\ |
| 1826 |
% \wordbox{1}{Total number of 16-bit data words that follow this |
| 1827 |
% header word, excluding the subsequent checksum-type value} \\ |
| 1828 |
% \bitbox{16}{Data~1} & \bitbox{16}{Data~2} \\ |
| 1829 |
% \bitbox{16}{Data~3} & \bitbox{16}{Data~4} \\ |
| 1830 |
% \bitbox[]{16}{$\vdots$ \\[1ex]} & |
| 1831 |
% \bitbox[]{16}{$\vdots$ \\[1ex]} \\ |
| 1832 |
% \bitbox{16}{Data~$N-1$} & \bitbox{16}{Data~$N$} \\ |
| 1833 |
% \bitbox{20}{\[ \mbox{A5A5}_{\mbox{\scriptsize H}} \oplus |
| 1834 |
% \left(\sum_{i=1}^N \mbox{Data}_i \right) \bmod 2^{20} \]} & |
| 1835 |
% \bitboxes*{1}{000010 000110} \\ |
| 1836 |
% \wordbox{2}{64-bit random number} |
| 1837 |
% \end{bytefield} |
| 1838 |
% \end{verbatim} |
| 1839 |
% |
| 1840 |
% \noindent |
| 1841 |
% Figure~\ref{fig:complex-diagram} shows the resulting protocol diagram. |
| 1842 |
% |
| 1843 |
% \begin{figure}[phtb] |
| 1844 |
% \centering |
| 1845 |
% \begin{lrbox}{\protocoldiagram} |
| 1846 |
% \begin{bytefield}[bitheight=2.5\baselineskip]{32} |
| 1847 |
% \bitheader{0,7,8,15,16,23,24,31} \\ |
| 1848 |
% \begin{rightwordgroup}{\parbox{6em}{\raggedright These words were taken |
| 1849 |
% verbatim from the TCP header definition (RFC~793).}} |
| 1850 |
% \bitbox{4}{Data offset} & \bitbox{6}{Reserved} & |
| 1851 |
% \bitbox{1}{\tiny U\\R\\G} & \bitbox{1}{\tiny A\\C\\K} & |
| 1852 |
% \bitbox{1}{\tiny P\\S\\H} & \bitbox{1}{\tiny R\\S\\T} & |
| 1853 |
% \bitbox{1}{\tiny S\\Y\\N} & \bitbox{1}{\tiny F\\I\\N} & |
| 1854 |
% \bitbox{16}{Window} \\ |
| 1855 |
% \bitbox{16}{Checksum} & \bitbox{16}{Urgent pointer} |
| 1856 |
% \end{rightwordgroup} \\ |
| 1857 |
% \wordbox[lrt]{1}{Data octets} \\ |
| 1858 |
% \skippedwords \\ |
| 1859 |
% \wordbox[lrb]{1}{} \\ |
| 1860 |
% \begin{leftwordgroup}{\parbox{6em}{\raggedright Note that we can display, |
| 1861 |
% for example, a misaligned 64-bit value with clever use of the |
| 1862 |
% optional argument to \texttt{\string\wordbox} and |
| 1863 |
% \texttt{\string\bitbox}.}} |
| 1864 |
% \bitbox{8}{Source} & \bitbox{8}{Destination} & \bitbox[lrt]{16}{} \\ |
| 1865 |
% \wordbox[lr]{1}{Timestamp} \\ |
| 1866 |
% \begin{rightwordgroup}{\parbox{6em}{\raggedright Why two Length fields? |
| 1867 |
% No particular reason.}} |
| 1868 |
% \bitbox[lrb]{16}{} & \bitbox{16}{Length} |
| 1869 |
% \end{leftwordgroup} \\ |
| 1870 |
% \bitbox{6}{Key} & \bitbox{6}{Value} & \bitbox{4}{Unused} & |
| 1871 |
% \bitbox{16}{Length} |
| 1872 |
% \end{rightwordgroup} \\ |
| 1873 |
% \wordbox{1}{Total number of 16-bit data words that follow this |
| 1874 |
% header word, excluding the subsequent checksum-type value} \\ |
| 1875 |
% \bitbox{16}{Data~1} & \bitbox{16}{Data~2} \\ |
| 1876 |
% \bitbox{16}{Data~3} & \bitbox{16}{Data~4} \\ |
| 1877 |
% \bitbox[]{16}{$\vdots$ \\[1ex]} & \bitbox[]{16}{$\vdots$ \\[1ex]} \\ |
| 1878 |
% \bitbox{16}{Data~$N-1$} & \bitbox{16}{Data~$N$} \\ |
| 1879 |
% \bitbox{20}{\[ \mbox{A5A5}_{\mbox{\scriptsize H}} \oplus |
| 1880 |
% \left(\sum_{i=1}^N \mbox{Data}_i \right) \bmod 2^{20} \]} & |
| 1881 |
% \bitboxes*{1}{000010 000110} \\ |
| 1882 |
% \wordbox{2}{64-bit random number} |
| 1883 |
% \end{bytefield} |
| 1884 |
% \end{lrbox} |
| 1885 |
% \makebox[0pt][c]{\usebox{\protocoldiagram}} |
| 1886 |
% \caption{Complex protocol diagram drawn with the \pkgname{bytefield} |
| 1887 |
% package} |
| 1888 |
% \label{fig:complex-diagram} |
| 1889 |
% \end{figure} |
| 1890 |
% |
| 1891 |
% |
| 1892 |
% \subsection{Upgrading from older versions} |
| 1893 |
% \label{sec:upgrading} |
| 1894 |
% |
| 1895 |
% \pkgname{bytefield}'s user interface changed substantially with the |
| 1896 |
% introduction of version~2.0. Because documents written for |
| 1897 |
% \pkgname{bytefield}~v1.\textit{x} will not build properly under later |
| 1898 |
% versions of the package, this section explains how to convert documents |
| 1899 |
% to the new interface. |
| 1900 |
% |
| 1901 |
% \begin{decl} |
| 1902 |
% \SpecialUsageIndex{\wordgroupr} |
| 1903 |
% |\wordgroupr| \\ |
| 1904 |
% \SpecialUsageIndex{\endwordgroupr} |
| 1905 |
% |\endwordgroupr| |
| 1906 |
% \end{decl} |
| 1907 |
% |
| 1908 |
% These have been replaced with the |rightwordgroup| environment to make |
| 1909 |
% their invocation more \LaTeX-like. Use |\begin{rightwordgroup}| |
| 1910 |
% instead of |\wordgroupr| and |\end{rightwordgroup}| instead of |
| 1911 |
% |\endwordgroupr|. |
| 1912 |
% |
| 1913 |
% \begin{decl} |
| 1914 |
% \SpecialUsageIndex{\wordgroupl} |
| 1915 |
% |\wordgroupl| \\ |
| 1916 |
% \SpecialUsageIndex{\endwordgroupl} |
| 1917 |
% |\endwordgroupl| |
| 1918 |
% \end{decl} |
| 1919 |
% |
| 1920 |
% These have been replaced with the |leftwordgroup| environment to make |
| 1921 |
% their invocation more \LaTeX-like. Use |\begin{leftwordgroup}| |
| 1922 |
% instead of |\wordgroupl| and |\end{leftwordgroup}| instead of |
| 1923 |
% |\endwordgroupl|. |
| 1924 |
% |
| 1925 |
% \begin{decl} |
| 1926 |
% \SpecialUsageIndex{\bitwidth} |
| 1927 |
% |\bitwidth| |
| 1928 |
% \end{decl} |
| 1929 |
% |
| 1930 |
% Instead of changing bit widths with |
| 1931 |
% |\setlength{\bitwidth}{|\meta{width}|}|, use |
| 1932 |
% |\bytefieldsetup{bitwidth=|\meta{width}|}|. |
| 1933 |
% |
| 1934 |
% \begin{decl} |
| 1935 |
% \SpecialUsageIndex{\byteheight} |
| 1936 |
% |\byteheight| |
| 1937 |
% \end{decl} |
| 1938 |
% |
| 1939 |
% Instead of changing bit heights with |
| 1940 |
% |\setlength{\byteheight}{|\meta{height}|}|, use |
| 1941 |
% |\bytefieldsetup{bitheight=|\meta{height}|}| (and note the change from |
| 1942 |
% ``|byte|'' to ``|bit|'' for consistency with \optname{bitwidth}). |
| 1943 |
% |
| 1944 |
% \begin{decl} |
| 1945 |
% \SpecialUsageIndex{\curlyspace} |
| 1946 |
% |\curlyspace| \\ |
| 1947 |
% \SpecialUsageIndex{\labelspace} |
| 1948 |
% |\labelspace| |
| 1949 |
% \end{decl} |
| 1950 |
% |
| 1951 |
% Instead of using |\setlength{\curlyspace}{|\meta{dist}|}| and |
| 1952 |
% |\setlength{\labelspace}{|\meta{dist}|}| to alter the horizontal space |
| 1953 |
% that appears before and after a curly brace, use |
| 1954 |
% |\bytefieldsetup{curlyspace=|\meta{dist}|}| and |
| 1955 |
% |\bytefieldsetup{labelspace=|\meta{dist}|}|. Note that, as described |
| 1956 |
% in Section~\ref{sec:basic-cmds}, left and right spacing can be set |
| 1957 |
% independently if desired. |
| 1958 |
% |
| 1959 |
% \begin{decl} |
| 1960 |
% \SpecialUsageIndex{\curlyshrinkage} |
| 1961 |
% |\curlyshrinkage| |
| 1962 |
% \end{decl} |
| 1963 |
% |
| 1964 |
% Instead of using |\setlength{\curlyshrinkage}{|\meta{dist}|}| to |
| 1965 |
% reduce the vertical space occupied by a curly brace, use |
| 1966 |
% |\bytefieldsetup{curlyshrinkage=|\meta{dist}|}|. Note that, as |
| 1967 |
% described in Section~\ref{sec:basic-cmds}, left and right curly-brace |
| 1968 |
% height can be reduced independently if desired. |
| 1969 |
% |
| 1970 |
% \begin{decl} |
| 1971 |
% \SpecialUsageIndex{\bitwidth} |
| 1972 |
% |\bitwidth| \oarg{endianness} \marg{bit-positions} |
| 1973 |
% \end{decl} |
| 1974 |
% |
| 1975 |
% The meaning of |\bitwidth|'s optional argument changed with |
| 1976 |
% \pkgname{bytefield}~v2.1. In older versions of the package, the |
| 1977 |
% optional argument was one of ``|l|'' or ``|b|'' for, respectively, |
| 1978 |
% little-endian or big-endian bit ordering. Starting with version~2.1, |
| 1979 |
% the optional argument can be any of the parameters described in |
| 1980 |
% Section~\ref{sec:options} (but practically only |bitformatting|, |
| 1981 |
% |endianness|, and |lsb|). Hence, ``|l|'' should be replaced with |
| 1982 |
% |endianness=little| and ``|b|'' should be replaced with |
| 1983 |
% |endianness=big|. Although more verbose, these new options can be |
| 1984 |
% specified once for the entire document by listing them as package |
| 1985 |
% options or as arguments to |\bytefieldsetup|. |
| 1986 |
% |
| 1987 |
% \bigskip |
| 1988 |
% |
| 1989 |
% As a crutch to help build older documents with minimal modification, |
| 1990 |
% \pkgname{bytefield} provides a |compat1| package option that restores |
| 1991 |
% the old interface. This option, invoked with |
| 1992 |
% |\usepackage[compat1]{bytefield}|, may disappear in a future version |
| 1993 |
% of the package and should therefore not be relied upon as a long-term |
| 1994 |
% approach to using \pkgname{bytefield}. |
| 1995 |
% |
| 1996 |
% |
| 1997 |
% \StopEventually{^^A |
| 1998 |
% |
| 1999 |
% \section{Future work} |
| 2000 |
% |
| 2001 |
% \pkgname{bytefield} is my first \LaTeX\ package, and, as such, there |
| 2002 |
% are a number of macros that could probably have been implemented a |
| 2003 |
% lot better. For example, \pkgname{bytefield} is somewhat wasteful |
| 2004 |
% of \meta{dimen} registers (although it did get a lot better with |
| 2005 |
% version~1.1 and again with version~1.3). The package should really |
| 2006 |
% get a major overhaul now that I've gotten better at \TeX\slash |
| 2007 |
% \LaTeX\ programming. One minor improvement I'd like to make in the |
| 2008 |
% package is to move left, small curly braces closer to the bit field. |
| 2009 |
% In the following figure, notice how distant the small curly appears |
| 2010 |
% from the bit-field body: |
| 2011 |
% |
| 2012 |
% \begin{bffigure} |
| 2013 |
% \bytefieldsetup{bitheight=4ex} |
| 2014 |
% \begin{bytefield}{16} |
| 2015 |
% \begin{leftwordgroup}{Too distant} |
| 2016 |
% \wordbox{1}{Something} |
| 2017 |
% \end{leftwordgroup} \\ |
| 2018 |
% \begin{leftwordgroup}{Looks okay} |
| 2019 |
% \wordbox{4}{Something else} |
| 2020 |
% \end{leftwordgroup} |
| 2021 |
% \end{bytefield} |
| 2022 |
% \end{bffigure} |
| 2023 |
% |
| 2024 |
% \noindent |
| 2025 |
% The problem is that the curly braces are left-aligned relative to |
| 2026 |
% each other, while they should be right-aligned. |
| 2027 |
% } |
| 2028 |
% |
| 2029 |
% |
| 2030 |
% \section{Implementation} |
| 2031 |
% |
| 2032 |
% \newcommand{\usermacro}{^^A |
| 2033 |
% \Needspace{2\baselineskip}^^A |
| 2034 |
% \marginpar{\huge\raisebox{-1.5ex}[0pt][1.5ex]{\strut\hspace{18pt}$\star$}\par}^^A |
| 2035 |
% } |
| 2036 |
% |
| 2037 |
% This section contains the complete source code for \pkgname{bytefield}. |
| 2038 |
% Most users will not get much out of it, but it should be of use to |
| 2039 |
% those who need more precise documentation and those who want to extend |
| 2040 |
% (or debug~\smiley) the \pkgname{bytefield} package. |
| 2041 |
% |
| 2042 |
% In this section, macros marked in the margin with a ``{\Large |
| 2043 |
% $\star$}'' are intended to be called by the user (and were described |
| 2044 |
% in Section~\ref{sec:usage}). All other macros are used only |
| 2045 |
% internally by \pkgname{bytefield}. |
| 2046 |
% |
| 2047 |
% \subsection{Required packages} |
| 2048 |
% |
| 2049 |
% Although |\widthof| and |\heightof| were introduced in June~1998, |
| 2050 |
% te\kern-1.5pt\TeX~2.0---still in widespread use at the time of this |
| 2051 |
% writing (2005)---ships with an earlier |calc.sty| in the |source| |
| 2052 |
% directory. Because a misconfigured system may find the |source| |
| 2053 |
% version of |calc.sty| we explicitly specify a later date when loading |
| 2054 |
% the |calc| package. |
| 2055 |
% \changes{v1.2a}{2005/07/31}{Specified an explicit package date when |
| 2056 |
% loading the \pkgname{calc} package to avoid loading an outdated |
| 2057 |
% version. Thanks to Kevin Quick for discovering that outdated versions |
| 2058 |
% of \pkgname{calc} are still being included in \TeX{} distributions.} |
| 2059 |
% \begin{macrocode} |
| 2060 |
\RequirePackage{calc}[1998/07/07] |
| 2061 |
\RequirePackage{keyval} |
| 2062 |
% \end{macrocode} |
| 2063 |
% |
| 2064 |
% |
| 2065 |
% \subsection{Utility macros} |
| 2066 |
% |
| 2067 |
% The following macros in this section are used by the box-drawing macros |
| 2068 |
% and the ``skipped words''-drawing macros. |
| 2069 |
% |
| 2070 |
% \begin{macro}{\bf@newdimen@old} |
| 2071 |
% \LaTeX's |\newdimen| defines new \meta{dimen}s globally. |
| 2072 |
% |\bf@newdimen@old| defines them locally. It simply merges |
| 2073 |
% \LaTeXe's |\newdimen| and |\alloc@| macros while omitting |\alloc@|'s |
| 2074 |
% ``|\global|'' declaration. The ``|old|'' in the name implies that |
| 2075 |
% |\bf@newdimen@old| is intended for use with older versions of \LaTeX. |
| 2076 |
% \begin{macrocode} |
| 2077 |
\def\bf@newdimen@old#1{\advance\count11 by 1 |
| 2078 |
\ch@ck1\insc@unt\dimen |
| 2079 |
\allocationnumber=\count11 |
| 2080 |
\dimendef#1=\allocationnumber |
| 2081 |
\wlog{\string#1=\string\dimen\the\allocationnumber\space (locally)}% |
| 2082 |
} |
| 2083 |
% \end{macrocode} |
| 2084 |
% \end{macro} |
| 2085 |
% |
| 2086 |
% \begin{macro}{\bf@e@alloc} |
| 2087 |
% Since 2014, \LaTeX\ has provided an internal |\e@alloc| macro that is |
| 2088 |
% used to define |\newcount|, |\newdimen|, |\newskip|, etc. |\bf@e@alloc| |
| 2089 |
% is a version of |\e@alloc| modified to allocate only within the current |
| 2090 |
% scope. (It elides all uses of |\global|.) We use |\bf@e@alloc| to |
| 2091 |
% define |\bf@newdimen|, which allocates a locally scoped \meta{dimen}. |
| 2092 |
% \begin{macrocode} |
| 2093 |
\def\bf@e@alloc#1#2#3#4#5#6{% |
| 2094 |
\advance#3\@ne |
| 2095 |
\e@ch@ck{#3}{#4}{#5}#1% |
| 2096 |
\allocationnumber#3\relax |
| 2097 |
#2#6\allocationnumber |
| 2098 |
\wlog{\string#6=\string#1\the\allocationnumber\space (locally)}% |
| 2099 |
} |
| 2100 |
% \end{macrocode} |
| 2101 |
% \end{macro} |
| 2102 |
% |
| 2103 |
% \begin{macro}{\bf@newdimen} |
| 2104 |
% \changes{v1.1}{2002/09/15}{Bug fix: Added \cs{bf@newdimen} to greatly |
| 2105 |
% reduce the likelihood of ``\texttt{No room for a new \string\cs{dimen}}'' |
| 2106 |
% errors (reported by Vitaly A. Repin)} |
| 2107 |
% \changes{v1.3}{2010/10/31}{Added support for $\varepsilon$-\TeX's larger |
| 2108 |
% local \meta{dimen} pool (code provided by Heiko Oberdiek)} |
| 2109 |
% \changes{v2.3}{2015/10/28}{Rewrote the macro based on discussions with |
| 2110 |
% David Carlisle to avoid producing ``\texttt{No room for a new |
| 2111 |
% \cs{dimen}}'' errors in newer versions of $\varepsilon$-\TeX\ |
| 2112 |
% (see \url{http://tex.stackexchange.com/q/275042})} |
| 2113 |
% \changes{v2.9}{2025/01/25}{Rewrite \cs{bf@newdimen} yet again so as to |
| 2114 |
% allocate \meta{dimen}s locally in newer versions of \LaTeX. Thanks to |
| 2115 |
% Qian Jin for the bug report} |
| 2116 |
% \pkgname{bytefield} allocates a number of \meta{dimen}s, largely for |
| 2117 |
% positioning curly braces. Because \TeX\ supports only 255 |
| 2118 |
% \meta{dimen}s, these normally would run out quickly for documents |
| 2119 |
% containing many bit fields. $\varepsilon$-\TeX\ provides many more |
| 2120 |
% \meta{dimen}s that the original \TeX, but it still would be preferable |
| 2121 |
% not to consume them unnecessarily. |
| 2122 |
% |
| 2123 |
% This is where |\bf@newdimen| comes in. If the \pkgname{elocalloc} |
| 2124 |
% package is loaded, |\bf@newdimen| is bound to \pkgname{elocalloc}'s |
| 2125 |
% |\locdimen| macro. Otherwise, if |\e@alloc| is defined---this is the |
| 2126 |
% common case---then |\bf@newdimen| is defined in terms of |\bf@e@alloc| |
| 2127 |
% just like |\newdimen| is defined in terms of |\e@alloc|. As a last |
| 2128 |
% resort, |\bf@newdimen| is bound to |\bf@newdimen@old| for use with old |
| 2129 |
% versions of \LaTeX. |
| 2130 |
% \begin{macrocode} |
| 2131 |
\AtBeginDocument{% |
| 2132 |
% \end{macrocode} |
| 2133 |
% Is \pkgname{elocalloc} loaded? |
| 2134 |
% \begin{macrocode} |
| 2135 |
\expandafter\ifx\csname locdimen\endcsname\relax |
| 2136 |
% \end{macrocode} |
| 2137 |
% No. Are we running a reasonably recent \LaTeX? |
| 2138 |
% \begin{macrocode} |
| 2139 |
\expandafter\ifx\csname e@alloc\endcsname\relax |
| 2140 |
% \end{macrocode} |
| 2141 |
% No. Use |\bf@newdimen@old|, which was designed in terms of primitives |
| 2142 |
% found in older versions of \LaTeX. |
| 2143 |
% \begin{macrocode} |
| 2144 |
\global\let\bf@newdimen=\bf@newdimen@old |
| 2145 |
\else |
| 2146 |
% \end{macrocode} |
| 2147 |
% We are running a reasonably recent \LaTeX\@. Define |\bf@newdimen| |
| 2148 |
% in terms of |\bf@e@alloc|, which allocates \meta{dimen}s locally. |
| 2149 |
% \begin{macrocode} |
| 2150 |
\gdef\bf@newdimen{% |
| 2151 |
\bf@e@alloc\dimen \dimendef {\count11}\insc@unt\float@count |
| 2152 |
}% |
| 2153 |
\fi |
| 2154 |
\else |
| 2155 |
% \end{macrocode} |
| 2156 |
% \pkgname{elocalloc} is loaded. Define |\bf@newdimen| in terms of |
| 2157 |
% \pkgname{elocalloc}'s |\locdimen| macro. |
| 2158 |
% \begin{macrocode} |
| 2159 |
\let\bf@newdimen=\locdimen |
| 2160 |
\fi |
| 2161 |
} |
| 2162 |
% \end{macrocode} |
| 2163 |
% \end{macro} |
| 2164 |
% |
| 2165 |
% \begin{macro}{\bytefield@height} |
| 2166 |
% \begin{macro}{\ifcounting@words} |
| 2167 |
% When |\ifcounting@words| is \textsc{true}, add the height of the next |
| 2168 |
% \texttt{picture} environment to |\bytefield@height|. We set |
| 2169 |
% |\counting@wordstrue| at the beginning of each word, and |
| 2170 |
% |\counting@wordsfalse| after each |\bitbox|, |\wordbox|, or |\skippedwords| |
| 2171 |
% picture. |
| 2172 |
% \begin{macrocode} |
| 2173 |
\newlength{\bytefield@height} |
| 2174 |
\newif\ifcounting@words |
| 2175 |
% \end{macrocode} |
| 2176 |
% \end{macro} |
| 2177 |
% \end{macro} |
| 2178 |
% |
| 2179 |
% \begin{macro}{\inc@bytefield@height} |
| 2180 |
% We have to define a special macro to increment |\bytefield@height| |
| 2181 |
% because the \pkgname{calc} package's |\addtolength| macro doesn't seem |
| 2182 |
% to see the global value. So we |\setlength| a temporary (to get |
| 2183 |
% \pkgname{calc}'s nice infix features) and |\advance| |\bytefield@height| |
| 2184 |
% by that amount. |
| 2185 |
% \begin{macrocode} |
| 2186 |
\newlength{\bytefield@height@increment} |
| 2187 |
\DeclareRobustCommand{\inc@bytefield@height}[1]{% |
| 2188 |
\setlength{\bytefield@height@increment}{#1}% |
| 2189 |
\global\advance\bytefield@height by \bytefield@height@increment |
| 2190 |
} |
| 2191 |
% \end{macrocode} |
| 2192 |
% \end{macro} |
| 2193 |
% |
| 2194 |
% \subsection{Top-level environment} |
| 2195 |
% |
| 2196 |
% \begin{macro}{\entire@bytefield@picture} |
| 2197 |
% Declare a box for containing the entire bytefield. By storing |
| 2198 |
% everything in a box and then typesetting it later (at the |
| 2199 |
% |\end{bytefield}|), we can center the bit field, put a box around it, |
| 2200 |
% and do other operations on the entire figure. |
| 2201 |
% \begin{macrocode} |
| 2202 |
\newsavebox{\entire@bytefield@picture} |
| 2203 |
% \end{macrocode} |
| 2204 |
% \end{macro} |
| 2205 |
% |
| 2206 |
% \begin{environment}{bytefield} |
| 2207 |
% \usermacro |
| 2208 |
% \changes{v2.4}{2017/09/15}{Make the code resilient to changes in |
| 2209 |
% \texttt{\string\string\string\baselinestretch}. Thanks to |
| 2210 |
% Karst Koymans for the bug report} |
| 2211 |
% \begin{macro}{\bits@wide} |
| 2212 |
% \begin{macro}{\old@nl} |
| 2213 |
% \begin{macro}{\amp} |
| 2214 |
% The |bytefield| environment contains the layout of bits in a sequence |
| 2215 |
% of words. This is the main environment defined by the |
| 2216 |
% \pkgname{bytefield} package. The argument is the number of bits wide |
| 2217 |
% the bytefield should be. We turn |&| into a space character so the |
| 2218 |
% user can think of a \pkgname{bytefield} as being analogous to a |
| 2219 |
% \texttt{tabular} environment, even though we're really setting the |
| 2220 |
% bulk of the picture in a single column. (Row labels go in separate |
| 2221 |
% columns, however.) |
| 2222 |
% \begin{macrocode} |
| 2223 |
\newenvironment{bytefield}[2][]{% |
| 2224 |
\bf@bytefieldsetup{#1}% |
| 2225 |
\renewcommand{\baselinestretch}{}% |
| 2226 |
\selectfont |
| 2227 |
\def\bits@wide{#2}% |
| 2228 |
\let\old@nl=\\% |
| 2229 |
\let\amp=&% |
| 2230 |
\catcode`\&=10 |
| 2231 |
\openup -1pt |
| 2232 |
\setlength{\bytefield@height}{0pt}% |
| 2233 |
\setlength{\unitlength}{1pt}% |
| 2234 |
\global\counting@wordstrue |
| 2235 |
\begin{lrbox}{\entire@bytefield@picture}% |
| 2236 |
% \end{macrocode} |
| 2237 |
% \end{macro} |
| 2238 |
% \end{macro} |
| 2239 |
% \end{macro} |
| 2240 |
% \begin{macro}{\\} |
| 2241 |
% \usermacro |
| 2242 |
% We redefine |\\| within the |bytefield| environment to make it aware |
| 2243 |
% of curly braces that surround the protocol diagram. |
| 2244 |
% \changes{v2.1}{2011/06/12}{Augmented the definition |
| 2245 |
% of~\texttt{\string\string\string\\} to accept an optional argument, |
| 2246 |
% just like in a \texttt{tabular} environment} |
| 2247 |
% \begin{macrocode} |
| 2248 |
\renewcommand{\\}[1][0pt]{% |
| 2249 |
\unskip |
| 2250 |
\vspace{##1}% |
| 2251 |
\amp\show@wordlabelr\cr |
| 2252 |
\ignorespaces\global\counting@wordstrue\make@lspace\amp}% |
| 2253 |
% \end{macrocode} |
| 2254 |
% \end{macro} |
| 2255 |
% \begin{macrocode} |
| 2256 |
\vbox\bgroup\ialign\bgroup##\amp##\amp##\cr\amp |
| 2257 |
}{% |
| 2258 |
\unskip |
| 2259 |
\amp\show@wordlabelr\cr\egroup\egroup |
| 2260 |
\end{lrbox}% |
| 2261 |
\usebox{\entire@bytefield@picture}% |
| 2262 |
} |
| 2263 |
% \end{macrocode} |
| 2264 |
% \end{environment} |
| 2265 |
% |
| 2266 |
% \subsection{Box-drawing macros} |
| 2267 |
% |
| 2268 |
% \subsubsection{Drawing (proper)} |
| 2269 |
% |
| 2270 |
% \begin{macro}{\bf@bitformatting} |
| 2271 |
% Format a bit number in the bit header. |\bf@bitformatting| may be |
| 2272 |
% redefined to take either a single argument (\`a~la |\textbf|) or no |
| 2273 |
% argument (\`a~la |\small|). |
| 2274 |
% \changes{v1.4}{2011/01/15}{Introduced this macro at Steven R. King's |
| 2275 |
% request to enable users to alter the bit header's font size} |
| 2276 |
% \begin{macrocode} |
| 2277 |
\newcommand*{\bf@bitformatting}{\tiny} |
| 2278 |
% \end{macrocode} |
| 2279 |
% \end{macro} |
| 2280 |
% |
| 2281 |
% \begin{macro}{\bf@boxformatting} |
| 2282 |
% Format the text within a bit box or word box. |\bf@boxformatting| |
| 2283 |
% takes either a single argument (\`a~la |\textbf|) or no argument |
| 2284 |
% (\`a~la |\small|). The text that follows |\bf@boxformatting| is |
| 2285 |
% guaranteed to be a group that ends in |\par|, so if |
| 2286 |
% |\bf@boxformatting| accepts an argument, the macro should be defined |
| 2287 |
% with |\long| (e.g.,~with |\newcommand| but not with |\newcommand*|). |
| 2288 |
% \begin{macrocode} |
| 2289 |
\newcommand*{\bf@boxformatting}{\centering} |
| 2290 |
% \end{macrocode} |
| 2291 |
% \end{macro} |
| 2292 |
% |
| 2293 |
% \begin{macro}{\bf@bitwidth} |
| 2294 |
% Define the width of a single bit. Note that this is wide enough to |
| 2295 |
% display a two-digit number without it running into adjacent numbers. |
| 2296 |
% For larger words, be sure to |\setlength| this larger. |
| 2297 |
% \begin{macrocode} |
| 2298 |
\newlength{\bf@bitwidth} |
| 2299 |
\settowidth{\bf@bitwidth}{\bf@bitformatting{99i}} |
| 2300 |
% \end{macrocode} |
| 2301 |
% \end{macro} |
| 2302 |
% |
| 2303 |
% \begin{macro}{\bf@bitheight} |
| 2304 |
% This is the height of a single bit within the bit field. |
| 2305 |
% \begin{macrocode} |
| 2306 |
\newlength{\bf@bitheight} |
| 2307 |
\setlength{\bf@bitheight}{4ex} |
| 2308 |
% \end{macrocode} |
| 2309 |
% \end{macro} |
| 2310 |
% |
| 2311 |
% \begin{macro}{\units@wide} |
| 2312 |
% \begin{macro}{\units@tall} |
| 2313 |
% These are scratch variables for storing the width and height (in |
| 2314 |
% points) of the box we're about to draw. |
| 2315 |
% \begin{macrocode} |
| 2316 |
\newlength{\units@wide} |
| 2317 |
\newlength{\units@tall} |
| 2318 |
% \end{macrocode} |
| 2319 |
% \end{macro} |
| 2320 |
% \end{macro} |
| 2321 |
% |
| 2322 |
% \begin{macro}{\bf@call@box@cmd} |
| 2323 |
% \begin{macro}{\bf@call@box@func} |
| 2324 |
% Define any box-drawing macro that accepts the same set of four |
| 2325 |
% arguments. It takes as input the name of a macro that is defined with |
| 2326 |
% formal parameters |[#1]#2[#3]#4|. \cs{bf@call@box@cmd} then invokes |
| 2327 |
% that macro, passing it a set of lines to draw out of the set |
| 2328 |
% ``\texttt{lrtbLRTB}''~(|#1|), a number of bits or words~(|#2|), a list |
| 2329 |
% of key/value pairs~(|#3|), and arbitrary text to typeset~(|#4|). |
| 2330 |
% \begin{macrocode} |
| 2331 |
\newcommand*{\bf@call@box@cmd}[1]{% |
| 2332 |
\def\bf@call@box@func{#1}% |
| 2333 |
\bf@call@box@cmd@i |
| 2334 |
} |
| 2335 |
% \end{macrocode} |
| 2336 |
% \end{macro} |
| 2337 |
% \end{macro} |
| 2338 |
% |
| 2339 |
% \begin{macro}{\bf@call@box@cmd@i} |
| 2340 |
% \begin{macro}{\bf@call@box@arg@i} |
| 2341 |
% \begin{macro}{\bf@call@box@arg@ii} |
| 2342 |
% Store the set of lines and the bit/word count and invoke |
| 2343 |
% \cs{bf@call@box@cmd@ii}. |
| 2344 |
% \begin{macrocode} |
| 2345 |
\newcommand*{\bf@call@box@cmd@i}[2][lrtb]{% |
| 2346 |
\def\bf@call@box@arg@i{#1}% |
| 2347 |
\def\bf@call@box@arg@ii{#2}% |
| 2348 |
\bf@call@box@cmd@ii |
| 2349 |
} |
| 2350 |
% \end{macrocode} |
| 2351 |
% \end{macro} |
| 2352 |
% \end{macro} |
| 2353 |
% \end{macro} |
| 2354 |
% |
| 2355 |
% \begin{macro}{\bf@call@box@cmd@ii} |
| 2356 |
% \begin{macro}{\bf@call@box@arg@iii} |
| 2357 |
% \begin{macro}{\bf@call@box@arg@iv} |
| 2358 |
% Store the key/value parameters and the text to typeset then invoke the |
| 2359 |
% macro originally passed to \cs{bf@call@box@cmd}. |
| 2360 |
% \begin{macrocode} |
| 2361 |
\newcommand*{\bf@call@box@cmd@ii}[2][]{% |
| 2362 |
\def\bf@call@box@arg@iii{#1}% |
| 2363 |
\def\bf@call@box@arg@iv{#2}% |
| 2364 |
\bf@call@box@func |
| 2365 |
} |
| 2366 |
% \end{macrocode} |
| 2367 |
% \end{macro} |
| 2368 |
% \end{macro} |
| 2369 |
% \end{macro} |
| 2370 |
% |
| 2371 |
% \begin{macro}{\bitbox} |
| 2372 |
% \usermacro |
| 2373 |
% Put some text~(|#4|) in a box that's a given number of bits~(|#2|) |
| 2374 |
% wide and one byte tall. An optional argument~(|#1|) specifies which |
| 2375 |
% lines to draw---|[l]|eft, |[r]|ight, |[t]|op, and/or |[b]|ottom |
| 2376 |
% (default: |lrtb|). Additional drawing parameters can be provided via |
| 2377 |
% another optional argument~(|#3|). |
| 2378 |
% \begin{macrocode} |
| 2379 |
\DeclareRobustCommand{\bitbox}{\bf@call@box@cmd{\bf@bitbox}} |
| 2380 |
% \end{macrocode} |
| 2381 |
% \end{macro} |
| 2382 |
% |
| 2383 |
% \begin{macro}{\bf@bitbox} |
| 2384 |
% Implement all of the \cs{bitbox} logic. |
| 2385 |
% \begin{macrocode} |
| 2386 |
\def\bf@bitbox{% |
| 2387 |
\bgroup |
| 2388 |
\expandafter\bf@parse@bitbox@arg\expandafter{\bf@call@box@arg@i}% |
| 2389 |
\setlength{\units@wide}{\bf@bitwidth * \bf@call@box@arg@ii}% |
| 2390 |
\expandafter\bf@bytefieldsetup\expandafter{\bf@call@box@arg@iii}% |
| 2391 |
\@ifundefined{bf@bgcolor}{% |
| 2392 |
}{% |
| 2393 |
% \end{macrocode} |
| 2394 |
% If \optname{bgcolor} was specified, draw a colored rule of the full |
| 2395 |
% size of the box. |
| 2396 |
% \begin{macrocode} |
| 2397 |
\rlap{% |
| 2398 |
\draw@bit@picture{\strip@pt\units@wide}{\strip@pt\bf@bitheight}{% |
| 2399 |
\color{\bf@bgcolor}% |
| 2400 |
\rule{\width}{\height}% |
| 2401 |
}% |
| 2402 |
}% |
| 2403 |
}% |
| 2404 |
% \end{macrocode} |
| 2405 |
% Draw the user-provided text on top of the rule (if any). |
| 2406 |
% \begin{macrocode} |
| 2407 |
\draw@bit@picture{\strip@pt\units@wide}{\strip@pt\bf@bitheight}{% |
| 2408 |
\bf@call@box@arg@iv |
| 2409 |
}% |
| 2410 |
\egroup |
| 2411 |
\ignorespaces |
| 2412 |
} |
| 2413 |
% \end{macrocode} |
| 2414 |
% \end{macro} |
| 2415 |
% |
| 2416 |
% \begin{macro}{\wordbox} |
| 2417 |
% \usermacro |
| 2418 |
% Put some text~(|#4|) in a box that's a given number of bytes~(|#2|) |
| 2419 |
% tall and one word (|\bits@wide| bits) wide. An optional |
| 2420 |
% argument~(|#1|) specifies which lines to draw---|[l]|eft, |[r]|ight, |
| 2421 |
% |[t]|op, and/or |[b]|ottom (default: |lrtb|). Additional drawing |
| 2422 |
% parameters can be provided via another optional argument~(|#3|). |
| 2423 |
% \begin{macrocode} |
| 2424 |
\DeclareRobustCommand{\wordbox}{\bf@call@box@cmd{\bf@wordbox}} |
| 2425 |
% \end{macrocode} |
| 2426 |
% \end{macro} |
| 2427 |
% |
| 2428 |
% \begin{macro}{\bf@wordbox} |
| 2429 |
% Implement all of the \cs{wordbox} logic. |
| 2430 |
% \begin{macrocode} |
| 2431 |
\def\bf@wordbox{% |
| 2432 |
\bgroup |
| 2433 |
\expandafter\bf@parse@bitbox@arg\expandafter{\bf@call@box@arg@i}% |
| 2434 |
\setlength{\units@wide}{\bf@bitwidth * \bits@wide}% |
| 2435 |
\setlength{\units@tall}{\bf@bitheight * \bf@call@box@arg@ii}% |
| 2436 |
\expandafter\bf@bytefieldsetup\expandafter{\bf@call@box@arg@iii}% |
| 2437 |
\@ifundefined{bf@bgcolor}{% |
| 2438 |
}{% |
| 2439 |
% \end{macrocode} |
| 2440 |
% If \optname{bgcolor} was specified, draw a colored rule of the full |
| 2441 |
% size of the box. |
| 2442 |
% \begin{macrocode} |
| 2443 |
\rlap{% |
| 2444 |
\draw@bit@picture{\strip@pt\units@wide}{\strip@pt\units@tall}{% |
| 2445 |
\color{\bf@bgcolor}% |
| 2446 |
\rule{\width}{\height}% |
| 2447 |
}% |
| 2448 |
}% |
| 2449 |
}% |
| 2450 |
% \end{macrocode} |
| 2451 |
% Draw the user-provided text on top of the rule (if any). |
| 2452 |
% \begin{macrocode} |
| 2453 |
\draw@bit@picture{\strip@pt\units@wide}{\strip@pt\units@tall}{% |
| 2454 |
\bf@call@box@arg@iv |
| 2455 |
}% |
| 2456 |
% \end{macrocode} |
| 2457 |
% Invoke the user-provided \cs{bf@per@word} macro once per word. |
| 2458 |
% \begin{macrocode} |
| 2459 |
\@ifundefined{bf@per@word}{}{\bf@invoke@per@word{\bf@call@box@arg@ii}}% |
| 2460 |
\egroup |
| 2461 |
\ignorespaces |
| 2462 |
} |
| 2463 |
% \end{macrocode} |
| 2464 |
% \end{macro} |
| 2465 |
% |
| 2466 |
% \begin{macro}{\draw@bit@picture} |
| 2467 |
% Put some text (|#3|) in a box that's a given number of units (|#1|) |
| 2468 |
% wide and a given number of units (|#2|) tall. |
| 2469 |
% We format the text with a |\parbox| to enable word-wrapping and explicit |
| 2470 |
% line breaks. In addition, we define |\height|, |\depth|, |\totalheight|, |
| 2471 |
% and |\width| (\`a la |\makebox| and friends), so the user can utilize |
| 2472 |
% those for special effects (e.g., a |\rule| that fills the entire box). |
| 2473 |
% As an added bonus, we define |\widthunits| and |\heightunits|, which |
| 2474 |
% are the width and height of the box in multiples of |\unitlength| |
| 2475 |
% (i.e., |#1| and |#2|, respectively). |
| 2476 |
% \begin{macrocode} |
| 2477 |
\DeclareRobustCommand{\draw@bit@picture}[3]{% |
| 2478 |
\begin{picture}(#1,#2)% |
| 2479 |
% \end{macrocode} |
| 2480 |
% First, we plot the user's text, with all sorts of useful lengths |
| 2481 |
% predefined. |
| 2482 |
% \begin{macrocode} |
| 2483 |
\put(0,0){\makebox(#1,#2){\parbox{#1\unitlength}{% |
| 2484 |
\bf@set@user@dimens{#1}{#2}% |
| 2485 |
\bf@boxformatting{#3\par}}}}% |
| 2486 |
% \end{macrocode} |
| 2487 |
% Next, we draw each line individually. I suppose we could make a special |
| 2488 |
% case for ``all lines'' and use a |\framebox| above, but the following |
| 2489 |
% works just fine. |
| 2490 |
% \begin{macrocode} |
| 2491 |
\ifbitbox@top |
| 2492 |
\put(0,#2){\line(1,0){#1}}% |
| 2493 |
\fi |
| 2494 |
\ifbitbox@bottom |
| 2495 |
\put(0,0){\line(1,0){#1}}% |
| 2496 |
\fi |
| 2497 |
\ifbitbox@left |
| 2498 |
\put(0,0){\line(0,1){#2}}% |
| 2499 |
\fi |
| 2500 |
\ifbitbox@right |
| 2501 |
\put(#1,0){\line(0,1){#2}}% |
| 2502 |
\fi |
| 2503 |
\end{picture}% |
| 2504 |
% \end{macrocode} |
| 2505 |
% Finally, we indicate that we're no longer at the beginning of a word. |
| 2506 |
% The following code structure (albeit with different arguments to |
| 2507 |
% |\inc@bytefield@height|) is repeated in various places throughout this |
| 2508 |
% package. We document it only here, however. |
| 2509 |
% \begin{macrocode} |
| 2510 |
\ifcounting@words |
| 2511 |
\inc@bytefield@height{\unitlength * \real{#2}}% |
| 2512 |
\global\counting@wordsfalse |
| 2513 |
\fi |
| 2514 |
} |
| 2515 |
% \end{macrocode} |
| 2516 |
% \end{macro} |
| 2517 |
% |
| 2518 |
% \begin{macro}{\bf@invoke@per@word} |
| 2519 |
% Invoke \cs{bf@per@word} once per word, passing it the (0-indexed) word |
| 2520 |
% number and total number of words. |
| 2521 |
% \begin{macrocode} |
| 2522 |
\newcommand{\bf@invoke@per@word}[1]{% |
| 2523 |
\begin{picture}(0,0)% |
| 2524 |
\@tempcnta=0 |
| 2525 |
\@tempdima=#1\bf@bitheight |
| 2526 |
% \end{macrocode} |
| 2527 |
% Make various useful dimensions available to \cs{bf@per@word}. |
| 2528 |
% \begin{macrocode} |
| 2529 |
\bf@set@user@dimens{\strip@pt\units@wide}{\strip@pt\units@tall}% |
| 2530 |
\loop |
| 2531 |
\advance\@tempdima by -\bf@bitheight |
| 2532 |
\bgroup |
| 2533 |
\put(-\strip@pt\units@wide, \strip@pt\@tempdima){% |
| 2534 |
\expandafter\bf@per@word\expandafter{\the\@tempcnta}{#1}% |
| 2535 |
}% |
| 2536 |
\egroup |
| 2537 |
\advance\@tempcnta by 1\relax |
| 2538 |
\ifnum#1>\@tempcnta |
| 2539 |
\repeat |
| 2540 |
\end{picture}% |
| 2541 |
} |
| 2542 |
% \end{macrocode} |
| 2543 |
% \end{macro} |
| 2544 |
% |
| 2545 |
% \begin{macro}{\bf@set@user@dimens} |
| 2546 |
% \mbox{}\vspace*{-1ex} |
| 2547 |
% \begin{macro}{\width} |
| 2548 |
% \usermacro |
| 2549 |
% \begin{macro}{\height} |
| 2550 |
% \usermacro |
| 2551 |
% \begin{macro}{\depth} |
| 2552 |
% \usermacro |
| 2553 |
% \begin{macro}{\totalheight} |
| 2554 |
% \usermacro |
| 2555 |
% \begin{macro}{\widthunits} |
| 2556 |
% \usermacro |
| 2557 |
% \begin{macro}{\heightunits} |
| 2558 |
% \usermacro |
| 2559 |
% Given a width in bits~(|#1|) and a height in words~(|#2|), make a |
| 2560 |
% number of box dimensions available to the author: \cs{width}, |
| 2561 |
% \cs{height}, \cs{depth}, \cs{totalheight}. Additionally, make the |
| 2562 |
% arguments available to the author via the \cs{widthunits} and |
| 2563 |
% \cs{heightunits} macros. |
| 2564 |
% \begin{macrocode} |
| 2565 |
\newcommand{\bf@set@user@dimens}[2]{% |
| 2566 |
\bf@newdimen\width |
| 2567 |
\bf@newdimen\height |
| 2568 |
\bf@newdimen\depth |
| 2569 |
\bf@newdimen\totalheight |
| 2570 |
\width=#1\unitlength |
| 2571 |
\height=#2\unitlength |
| 2572 |
\depth=0pt% |
| 2573 |
\totalheight=#2\unitlength |
| 2574 |
\def\widthunits{#1}% |
| 2575 |
\def\heightunits{#2}% |
| 2576 |
} |
| 2577 |
% \end{macrocode} |
| 2578 |
% \end{macro} |
| 2579 |
% \end{macro} |
| 2580 |
% \end{macro} |
| 2581 |
% \end{macro} |
| 2582 |
% \end{macro} |
| 2583 |
% \end{macro} |
| 2584 |
% \end{macro} |
| 2585 |
% |
| 2586 |
% \begin{macro}{\bitboxes} |
| 2587 |
% \usermacro |
| 2588 |
% \changes{v2.2}{2014/06/01}{Added this macro based on an idea proposed |
| 2589 |
% by Andrew Mertz} |
| 2590 |
% \begin{macro}{\bitboxes*} |
| 2591 |
% \usermacro |
| 2592 |
% Put each token in |#3| into a box that's a given number of bits (|#2|) |
| 2593 |
% wide and one byte tall. An optional argument (|#1|) specifies which |
| 2594 |
% lines to draw---|[l]|eft, |[r]|ight, |[t]|op, and/or |[b]|ottom |
| 2595 |
% (default: |lrtb|). The |*|-form of the command omits interior left |
| 2596 |
% and right lines. |
| 2597 |
% \begin{macrocode} |
| 2598 |
\DeclareRobustCommand{\bitboxes}{% |
| 2599 |
\@ifstar |
| 2600 |
{\bf@call@box@cmd{\bf@bitboxes@star}}% |
| 2601 |
{\bf@call@box@cmd{\bf@bitboxes@no@star}}% |
| 2602 |
} |
| 2603 |
% \end{macrocode} |
| 2604 |
% \end{macro} |
| 2605 |
% \end{macro} |
| 2606 |
% |
| 2607 |
% \begin{macro}{\bf@relax} |
| 2608 |
% Define a macro that expands to |\relax| for use with |\ifx| tests |
| 2609 |
% against |\bf@bitboxes@arg|, which can contain either tokens to typeset |
| 2610 |
% or |\relax|. |
| 2611 |
% \begin{macrocode} |
| 2612 |
\def\bf@relax{\relax} |
| 2613 |
% \end{macrocode} |
| 2614 |
% \end{macro} |
| 2615 |
% |
| 2616 |
% \begin{macro}{\bf@bitboxes@no@star} |
| 2617 |
% Implement the unstarred version of \cs{bitboxes}. This macro simply |
| 2618 |
% expands its text argument into a list of tokens followed by \cs{relax} |
| 2619 |
% then invokes \cs{bf@bitboxes@no@star@i}. |
| 2620 |
% \begin{macrocode} |
| 2621 |
\def\bf@bitboxes@no@star{% |
| 2622 |
\expandafter\bf@bitboxes@no@star@i\bf@call@box@arg@iv\relax |
| 2623 |
\ignorespaces |
| 2624 |
} |
| 2625 |
% \end{macrocode} |
| 2626 |
% \end{macro} |
| 2627 |
% |
| 2628 |
% \begin{macro}{\bf@bitboxes@no@star@i} |
| 2629 |
% Walk the subsequent tokens one-by-one until \cs{relax} is encountered. |
| 2630 |
% For each token, invoke \cs{bf@bitbox} (the internal version of |
| 2631 |
% \cs{bitbox} for which \cs{bf@call@box@arg@}\meta{number} are all |
| 2632 |
% defined. |
| 2633 |
% \begin{macrocode} |
| 2634 |
\def\bf@bitboxes@no@star@i#1{% |
| 2635 |
\def\bf@call@box@arg@iv{#1}% |
| 2636 |
\ifx\bf@call@box@arg@iv\bf@relax |
| 2637 |
\let\next=\relax |
| 2638 |
\else |
| 2639 |
\bf@bitbox |
| 2640 |
\let\next=\bf@bitboxes@no@star@i |
| 2641 |
\fi |
| 2642 |
\next |
| 2643 |
} |
| 2644 |
% \end{macrocode} |
| 2645 |
% \end{macro} |
| 2646 |
% |
| 2647 |
% \begin{macro}{\bf@bitboxes@star} |
| 2648 |
% \begin{macro}{\bf@bitboxes@sides} |
| 2649 |
% Implement the starred version of \cs{bitboxes}. This macro simply |
| 2650 |
% stores the original \meta{sides} argument in \cs{bf@bitboxes@sides}, |
| 2651 |
% expands its text argument into a list of tokens followed by two |
| 2652 |
% \cs{relax}es, and invokes \cs{bf@bitboxes@star@i}. |
| 2653 |
% \begin{macrocode} |
| 2654 |
\def\bf@bitboxes@star{% |
| 2655 |
\edef\bf@bitboxes@sides{\bf@call@box@arg@i}% |
| 2656 |
\expandafter\bf@bitboxes@star@i\bf@call@box@arg@iv\relax\relax |
| 2657 |
\ignorespaces |
| 2658 |
} |
| 2659 |
% \end{macrocode} |
| 2660 |
% \end{macro} |
| 2661 |
% \end{macro} |
| 2662 |
% |
| 2663 |
% \begin{macro}{\bf@bitboxes@star@i} |
| 2664 |
% \begin{macro}{\bf@call@box@arg@iv} |
| 2665 |
% \begin{macro}{\bf@bitboxes@arg@ii} |
| 2666 |
% \begin{macro}{\next} |
| 2667 |
% Process the first token in the text argument passed to \cs{bitboxes*}. |
| 2668 |
% If it's also the last token (indicated by its being followed by |
| 2669 |
% \cs{relax}), draw an ordinary bit box with all sides present. If it's |
| 2670 |
% not the last token, draw a bit box with the right side suppressed and |
| 2671 |
% invoke \cs{bf@bitboxes@star@ii} on the remaining tokens. |
| 2672 |
% \begin{macrocode} |
| 2673 |
\def\bf@bitboxes@star@i#1#2{% |
| 2674 |
\def\bf@call@box@arg@iv{#1}% |
| 2675 |
\def\bf@bitboxes@arg@ii{#2}% |
| 2676 |
\ifx\bf@bitboxes@arg@ii\bf@relax |
| 2677 |
\bf@bitbox |
| 2678 |
\let\next=\relax |
| 2679 |
\else |
| 2680 |
\edef\bf@call@box@arg@i{\bf@bitboxes@sides R}% |
| 2681 |
\bf@bitbox |
| 2682 |
\def\next{\bf@bitboxes@star@ii{#2}}% |
| 2683 |
\fi |
| 2684 |
\next |
| 2685 |
} |
| 2686 |
% \end{macrocode} |
| 2687 |
% \end{macro} |
| 2688 |
% \end{macro} |
| 2689 |
% \end{macro} |
| 2690 |
% \end{macro} |
| 2691 |
% |
| 2692 |
% \begin{macro}{\bf@bitboxes@star@ii} |
| 2693 |
% \begin{macro}{\bf@call@box@arg@iv} |
| 2694 |
% \begin{macro}{\bf@bitboxes@arg@ii} |
| 2695 |
% \begin{macro}{\next} |
| 2696 |
% Process the second and subsequent tokens in the text argument passed |
| 2697 |
% to \cs{bitboxes*}. If the next token in the stream is the final one |
| 2698 |
% (indicated by its being followed by \cs{relax}), draw a bit box with |
| 2699 |
% the left side suppressed. If it's not the final token, draw a bit box |
| 2700 |
% with both the left and right sides suppressed and invoke itself |
| 2701 |
% recursively on the remaining tokens. |
| 2702 |
% \begin{macrocode} |
| 2703 |
\def\bf@bitboxes@star@ii#1#2{% |
| 2704 |
\def\bf@call@box@arg@iv{#1}% |
| 2705 |
\def\bf@bitboxes@arg@ii{#2}% |
| 2706 |
\ifx\bf@bitboxes@arg@ii\bf@relax |
| 2707 |
\edef\bf@call@box@arg@i{\bf@bitboxes@sides L}% |
| 2708 |
\else |
| 2709 |
\edef\bf@call@box@arg@i{\bf@bitboxes@sides LR}% |
| 2710 |
\fi |
| 2711 |
\ifx\bf@call@box@arg@iv\bf@relax |
| 2712 |
\let\next=\relax |
| 2713 |
\else |
| 2714 |
\bf@bitbox |
| 2715 |
\def\next{\bf@bitboxes@star@ii{#2}}% |
| 2716 |
\fi |
| 2717 |
\next |
| 2718 |
} |
| 2719 |
% \end{macrocode} |
| 2720 |
% \end{macro} |
| 2721 |
% \end{macro} |
| 2722 |
% \end{macro} |
| 2723 |
% \end{macro} |
| 2724 |
% |
| 2725 |
% |
| 2726 |
% \subsubsection{Parsing arguments} |
| 2727 |
% |
| 2728 |
% The macros in this section are used to parse the optional argument to |
| 2729 |
% \cs{bitbox}, \cs{wordbox}, and \cs{bitboxes}, which is some subset of |
| 2730 |
% $\{ |l|, |r|, |t|, |b|, |L|, |R|, |T|, |B| \}$ and defaults to |
| 2731 |
% ``|lrtb|'' for all three user macros. If the argument is empty, no |
| 2732 |
% lines are drawn. Lowercase letters in the argument display, |
| 2733 |
% respectively, the left, right, top, or bottom side of a box. |
| 2734 |
% Uppercase letters undo the effect of the corresponding, prior, |
| 2735 |
% lowercase letter and are used internally by \cs{bitboxes} to suppress |
| 2736 |
% internal left and right lines. |
| 2737 |
% |
| 2738 |
% \begin{macro}{\ifbitbox@top} |
| 2739 |
% \begin{macro}{\ifbitbox@bottom} |
| 2740 |
% \begin{macro}{\ifbitbox@left} |
| 2741 |
% \begin{macro}{\ifbitbox@right} |
| 2742 |
% These macros are set to \textsc{true} if we're to draw the |
| 2743 |
% corresponding edge on the subsequent |\bitbox| or |\wordbox|. |
| 2744 |
% \begin{macrocode} |
| 2745 |
\newif\ifbitbox@top |
| 2746 |
\newif\ifbitbox@bottom |
| 2747 |
\newif\ifbitbox@left |
| 2748 |
\newif\ifbitbox@right |
| 2749 |
% \end{macrocode} |
| 2750 |
% \end{macro} |
| 2751 |
% \end{macro} |
| 2752 |
% \end{macro} |
| 2753 |
% \end{macro} |
| 2754 |
% |
| 2755 |
% \begin{macro}{\bf@parse@bitbox@arg} |
| 2756 |
% This main parsing macro merely resets the above conditionals and |
| 2757 |
% calls a helper function, |\bf@parse@bitbox@sides|. |
| 2758 |
% \begin{macrocode} |
| 2759 |
\def\bf@parse@bitbox@arg#1{% |
| 2760 |
\bitbox@topfalse |
| 2761 |
\bitbox@bottomfalse |
| 2762 |
\bitbox@leftfalse |
| 2763 |
\bitbox@rightfalse |
| 2764 |
\bf@parse@bitbox@sides#1X% |
| 2765 |
} |
| 2766 |
% \end{macrocode} |
| 2767 |
% \end{macro} |
| 2768 |
% |
| 2769 |
% \begin{macro}{\bf@parse@bitbox@sides} |
| 2770 |
% The helper function for |\bf@parse@bitbox@arg| parses a single letter, |
| 2771 |
% sets the appropriate conditional to \textsc{true}, and calls itself |
| 2772 |
% tail-recursively until it sees an~``|X|''. |
| 2773 |
% \begin{macrocode} |
| 2774 |
\def\bf@parse@bitbox@sides#1{% |
| 2775 |
\ifx#1X% |
| 2776 |
\else |
| 2777 |
\ifx#1t% |
| 2778 |
\bitbox@toptrue |
| 2779 |
\else |
| 2780 |
\ifx#1b% |
| 2781 |
\bitbox@bottomtrue |
| 2782 |
\else |
| 2783 |
\ifx#1l% |
| 2784 |
\bitbox@lefttrue |
| 2785 |
\else |
| 2786 |
\ifx#1r% |
| 2787 |
\bitbox@righttrue |
| 2788 |
\else |
| 2789 |
\ifx#1T% |
| 2790 |
\bitbox@topfalse |
| 2791 |
\else |
| 2792 |
\ifx#1B% |
| 2793 |
\bitbox@bottomfalse |
| 2794 |
\else |
| 2795 |
\ifx#1L% |
| 2796 |
\bitbox@leftfalse |
| 2797 |
\else |
| 2798 |
\ifx#1R% |
| 2799 |
\bitbox@rightfalse |
| 2800 |
\else |
| 2801 |
\PackageWarning{bytefield}{Unrecognized box side `#1'}% |
| 2802 |
\fi |
| 2803 |
\fi |
| 2804 |
\fi |
| 2805 |
\fi |
| 2806 |
\fi |
| 2807 |
\fi |
| 2808 |
\fi |
| 2809 |
\fi |
| 2810 |
\expandafter\bf@parse@bitbox@sides |
| 2811 |
\fi |
| 2812 |
} |
| 2813 |
% \end{macrocode} |
| 2814 |
% \end{macro} |
| 2815 |
% |
| 2816 |
% \subsection{Skipped words} |
| 2817 |
% |
| 2818 |
% \begin{macro}{\units@high} |
| 2819 |
% This is the height of each diagonal line in the |\skippedwords| |
| 2820 |
% graphic. Note that |\units@high|~$=$ |\units@tall|~$-$ |
| 2821 |
% \textit{optional argument to} |\skippedwords|. |
| 2822 |
% \begin{macrocode} |
| 2823 |
\newlength{\units@high} |
| 2824 |
% \end{macrocode} |
| 2825 |
% \end{macro} |
| 2826 |
% |
| 2827 |
% \begin{macro}{\skippedwords} |
| 2828 |
% \usermacro |
| 2829 |
% Output a fancy graphic representing skipped words. The optional argument |
| 2830 |
% is the vertical space between the two diagonal lines (default: |2ex|). |
| 2831 |
% \begin{macrocode} |
| 2832 |
\DeclareRobustCommand{\skippedwords}[1][2ex]{% |
| 2833 |
\setlength{\units@wide}{\bf@bitwidth * \bits@wide}% |
| 2834 |
\setlength{\units@high}{1pt * \ratio{\units@wide}{6.0pt}}% |
| 2835 |
\setlength{\units@tall}{#1 + \units@high}% |
| 2836 |
\edef\num@wide{\strip@pt\units@wide}% |
| 2837 |
\edef\num@tall{\strip@pt\units@tall}% |
| 2838 |
\edef\num@high{\strip@pt\units@high}% |
| 2839 |
\begin{picture}(\num@wide,\num@tall) |
| 2840 |
\put(0,\num@tall){\line(6,-1){\num@wide}} |
| 2841 |
\put(\num@wide,0){\line(-6,1){\num@wide}} |
| 2842 |
\put(0,0){\line(0,1){\num@high}} |
| 2843 |
\put(\num@wide,\num@tall){\line(0,-1){\num@high}} |
| 2844 |
\end{picture}% |
| 2845 |
\ifcounting@words |
| 2846 |
\inc@bytefield@height{\unitlength * \real{\num@tall}}% |
| 2847 |
\global\counting@wordsfalse |
| 2848 |
\fi |
| 2849 |
} |
| 2850 |
% \end{macrocode} |
| 2851 |
% \end{macro} |
| 2852 |
% |
| 2853 |
% \subsection{Bit-position labels} |
| 2854 |
% |
| 2855 |
% \begin{macro}{\bf@bit@endianness} |
| 2856 |
% \pkgname{bytefield} can label bit headers in either little-endian ($0, |
| 2857 |
% 1, 2,~\ldots, N-1$) or big-endian ($N-1, N-2, N-3,~\ldots, 0$) fashion. |
| 2858 |
% The |\bf@bit@endianness| macro specifies which to use, either ``|l|'' |
| 2859 |
% for little-endian (the default) or ``|b|'' for big-endian. |
| 2860 |
% \begin{macrocode} |
| 2861 |
\newcommand*{\bf@bit@endianness}{l} |
| 2862 |
% \end{macrocode} |
| 2863 |
% \end{macro} |
| 2864 |
% |
| 2865 |
% \begin{macro}{\bf@first@bit} |
| 2866 |
% Normally, bits are numbered starting from zero. However, |
| 2867 |
% |\bf@first@bit| can be altered (usually locally) to begin numbering |
| 2868 |
% from a different value. |
| 2869 |
% \begin{macrocode} |
| 2870 |
\newcommand*{\bf@first@bit}{0} |
| 2871 |
% \end{macrocode} |
| 2872 |
% \end{macro} |
| 2873 |
% |
| 2874 |
% \begin{macro}{\bitheader} |
| 2875 |
% \usermacro |
| 2876 |
% Output a header of numbered bit positions. The optional argument (|#1|) |
| 2877 |
% is ``|l|'' for little-endian (default) or ``|b|'' for big-endian. |
| 2878 |
% The required argument (|#2|) is a list of bit positions to label. |
| 2879 |
% It is composed of comma-separated ranges of numbers, for example, |
| 2880 |
% ``|0-31|'', ``|0,7-8,15-16,23-24,31|'', or even something odd like |
| 2881 |
% ``|0-7,15-23|''. Ranges must be specified in increasing order; use the |
| 2882 |
% \optname{lsb} option to reverse the labels' direction. |
| 2883 |
% \changes{v2.1}{2011/06/17}{Changed the optional argument to accept |
| 2884 |
% \meta{key}!=\meta{value} pairs instead of just ``\texttt{l}'' and |
| 2885 |
% ``\texttt{b}''} |
| 2886 |
% \begin{macrocode} |
| 2887 |
\DeclareRobustCommand{\bitheader}[2][]{% |
| 2888 |
\bf@parse@bitbox@arg{lrtb}% |
| 2889 |
\setlength{\units@wide}{\bf@bitwidth * \bits@wide}% |
| 2890 |
\setlength{\units@tall}{\heightof{\bf@bitformatting{1234567890}}}% |
| 2891 |
\setlength{\units@high}{\units@tall * -1}% |
| 2892 |
\bf@process@bitheader@opts{#1}% |
| 2893 |
\begin{picture}(\strip@pt\units@wide,\strip@pt\units@tall)% |
| 2894 |
(0,\strip@pt\units@high) |
| 2895 |
\bf@parse@range@list#2,X, |
| 2896 |
\end{picture}% |
| 2897 |
\ifcounting@words |
| 2898 |
\inc@bytefield@height{\unitlength * \real{\strip@pt\units@tall}}% |
| 2899 |
\global\counting@wordsfalse |
| 2900 |
\fi |
| 2901 |
\ignorespaces |
| 2902 |
} |
| 2903 |
% \end{macrocode} |
| 2904 |
% \end{macro} |
| 2905 |
% |
| 2906 |
% \begin{macro}{\bf@parse@range@list} |
| 2907 |
% This is helper function~\#1 for |\bitheader|. It parses a |
| 2908 |
% comma-separated list of ranges, calling |\bf@parse@range| on each |
| 2909 |
% range. |
| 2910 |
% \changes{v1.1}{2002/06/24}{Bug fix: Swapped order of arguments to |
| 2911 |
% \texttt{\string\bslash\space ifx} test (suggested by Hans-Joachim |
| 2912 |
% Widmaier)} |
| 2913 |
% \begin{macrocode} |
| 2914 |
\def\bf@parse@range@list#1,{% |
| 2915 |
\ifx X#1 |
| 2916 |
\else |
| 2917 |
\bf@parse@range#1-#1-#1\relax |
| 2918 |
\expandafter\bf@parse@range@list |
| 2919 |
\fi |
| 2920 |
} |
| 2921 |
% \end{macrocode} |
| 2922 |
% \end{macro} |
| 2923 |
% |
| 2924 |
% \begin{macro}{\header@xpos} |
| 2925 |
% \begin{macro}{header@val} |
| 2926 |
% \begin{macro}{max@header@val} |
| 2927 |
% Define some miscellaneous variables to be used internally by |
| 2928 |
% |\bf@parse@range|: $x$~position of header, current label to output, |
| 2929 |
% and maximum label to output ($+ 1$). |
| 2930 |
% \begin{macrocode} |
| 2931 |
\newlength{\header@xpos} |
| 2932 |
\newcounter{header@val} |
| 2933 |
\newcounter{max@header@val} |
| 2934 |
% \end{macrocode} |
| 2935 |
% \end{macro} |
| 2936 |
% \end{macro} |
| 2937 |
% \end{macro} |
| 2938 |
% |
| 2939 |
% \begin{macro}{\bf@parse@range} |
| 2940 |
% This is helper function~\#2 for |\bitheader|. It parses a |
| 2941 |
% hyphen-separated pair of numbers (or a single number) and displays the |
| 2942 |
% number at the correct bit position. |
| 2943 |
% \changes{v2.1}{2011/06/16}{Added code due to Renaud Pacalet for shifting |
| 2944 |
% the bit header by a distance corresponding to |
| 2945 |
% \texttt{\string\string\string\bf@first@bit}, used for typesetting |
| 2946 |
% registers split across rows} |
| 2947 |
% \begin{macrocode} |
| 2948 |
\def\bf@parse@range#1-#2-#3\relax{% |
| 2949 |
\setcounter{header@val}{#1} |
| 2950 |
\setcounter{max@header@val}{#2 + 1} |
| 2951 |
\loop |
| 2952 |
\ifnum\value{header@val}<\value{max@header@val}% |
| 2953 |
\if\bf@bit@endianness b% |
| 2954 |
\setlength{\header@xpos}{% |
| 2955 |
\bf@bitwidth * (\bits@wide - \value{header@val} + \bf@first@bit - 1)}% |
| 2956 |
\else |
| 2957 |
\setlength{\header@xpos}{\bf@bitwidth * (\value{header@val} - \bf@first@bit)}% |
| 2958 |
\fi |
| 2959 |
\put(\strip@pt\header@xpos,0){% |
| 2960 |
\makebox(\strip@pt\bf@bitwidth,\strip@pt\units@tall){% |
| 2961 |
\bf@bitformatting{\theheader@val}}} |
| 2962 |
\addtocounter{header@val}{1} |
| 2963 |
\repeat |
| 2964 |
} |
| 2965 |
% \end{macrocode} |
| 2966 |
% \end{macro} |
| 2967 |
% |
| 2968 |
% \begin{macro}{\bf@process@bitheader@opts} |
| 2969 |
% \begin{macro}{\KV@bytefield@l} |
| 2970 |
% \begin{macro}{\KV@bytefield@b} |
| 2971 |
% \begin{macro}{\KV@bytefield@l@default} |
| 2972 |
% \begin{macro}{\KV@bytefield@b@default} |
| 2973 |
% This is helper function~\#3 for |\bitheader|. It processes the |
| 2974 |
% optional argument to |\bitheader|. |
| 2975 |
% \begin{macrocode} |
| 2976 |
\newcommand*{\bf@process@bitheader@opts}{% |
| 2977 |
\let\KV@bytefield@l=\KV@bitheader@l |
| 2978 |
\let\KV@bytefield@b=\KV@bitheader@b |
| 2979 |
\let\KV@bytefield@l@default=\KV@bitheader@l@default |
| 2980 |
\let\KV@bytefield@b@default=\KV@bitheader@b@default |
| 2981 |
\setkeys{bytefield}% |
| 2982 |
} |
| 2983 |
% \end{macrocode} |
| 2984 |
% \end{macro} |
| 2985 |
% \end{macro} |
| 2986 |
% \end{macro} |
| 2987 |
% \end{macro} |
| 2988 |
% \end{macro} |
| 2989 |
% |
| 2990 |
% \begin{macro}{\KV@bitheader@l} |
| 2991 |
% \begin{macro}{\KV@bitheader@b} |
| 2992 |
% For backwards compatibility we also accept the (now deprecated) |l| as |
| 2993 |
% a synonym for |endianness=little| and |b| as a synonym for |
| 2994 |
% |endianness=big|. A typical document will specify an |endianness| |
| 2995 |
% option not as an argument to |\bitheader| but rather as a package |
| 2996 |
% option that applies to the entire document. If the \optname{compat1} |
| 2997 |
% option was provided to \pkgname{bytefield} (determined below by the |
| 2998 |
% existence of the |\curlyshrinkage| control word), we suppress the |
| 2999 |
% deprecation warning message. |
| 3000 |
% \begin{macrocode} |
| 3001 |
\define@key{bitheader}{l}[true]{% |
| 3002 |
\expandafter\ifx\csname curlyshrinkage\endcsname\relax |
| 3003 |
\PackageWarning{bytefield}{% |
| 3004 |
The "l" argument to \protect\bitheader\space is deprecated.\MessageBreak |
| 3005 |
Instead, please use "endianness=little", which can\MessageBreak |
| 3006 |
even be declared globally for the entire document.\MessageBreak |
| 3007 |
This warning occurred}% |
| 3008 |
\fi |
| 3009 |
\def\bf@bit@endianness{l}% |
| 3010 |
} |
| 3011 |
\define@key{bitheader}{b}[true]{% |
| 3012 |
\expandafter\ifx\csname curlyshrinkage\endcsname\relax |
| 3013 |
\PackageWarning{bytefield}{% |
| 3014 |
The "b" argument to \protect\bitheader\space is deprecated.\MessageBreak |
| 3015 |
Instead, please use "endianness=big", which can\MessageBreak |
| 3016 |
even be declared globally for the entire document.\MessageBreak |
| 3017 |
This warning occurred}% |
| 3018 |
\fi |
| 3019 |
\def\bf@bit@endianness{b}% |
| 3020 |
} |
| 3021 |
% \end{macrocode} |
| 3022 |
% \end{macro} |
| 3023 |
% \end{macro} |
| 3024 |
% |
| 3025 |
% |
| 3026 |
% \subsection{Word labels} |
| 3027 |
% |
| 3028 |
% \subsubsection{Curly-brace manipulation} |
| 3029 |
% |
| 3030 |
% \begin{macro}{\bf@leftcurlyshrinkage} |
| 3031 |
% \begin{macro}{\bf@rightcurlyshrinkage} |
| 3032 |
% Reduce the height of a left (right) curly brace by |
| 3033 |
% |\bf@leftcurlyshrinkage| (|\bf@rightcurlyshrinkage|) so its ends don't |
| 3034 |
% overlap whatever is above or below it. The default value (5\,pt.) was |
| 3035 |
% determined empirically and shouldn't need to be changed. However, on |
| 3036 |
% the off-chance the user employs a math font with very different curly |
| 3037 |
% braces from Computer Modern's, |\bf@leftcurlyshrinkage| and |
| 3038 |
% |\bf@rightcurlyshrinkage| can be modified. |
| 3039 |
% \begin{macrocode} |
| 3040 |
\def\bf@leftcurlyshrinkage{5pt} |
| 3041 |
\def\bf@rightcurlyshrinkage{5pt} |
| 3042 |
% \end{macrocode} |
| 3043 |
% \end{macro} |
| 3044 |
% \end{macro} |
| 3045 |
% |
| 3046 |
% \begin{macro}{\bf@leftcurlyspace} |
| 3047 |
% \begin{macro}{\bf@rightcurlyspace} |
| 3048 |
% \begin{macro}{\bf@leftlabelspace} |
| 3049 |
% \begin{macro}{\bf@rightlabelspace} |
| 3050 |
% Define the amount of space to insert before a curly brace and before a |
| 3051 |
% word label (i.e.,~after a curly brace). |
| 3052 |
% \begin{macrocode} |
| 3053 |
\def\bf@leftcurlyspace{1ex} |
| 3054 |
\def\bf@rightcurlyspace{1ex} |
| 3055 |
\def\bf@leftlabelspace{0.5ex} |
| 3056 |
\def\bf@rightlabelspace{0.5ex} |
| 3057 |
% \end{macrocode} |
| 3058 |
% \end{macro} |
| 3059 |
% \end{macro} |
| 3060 |
% \end{macro} |
| 3061 |
% \end{macro} |
| 3062 |
% |
| 3063 |
% \begin{macro}{\bf@leftcurly} |
| 3064 |
% \begin{macro}{\bf@rightcurly} |
| 3065 |
% Define the symbols to use as left and right curly braces. These |
| 3066 |
% symbols must be extensible math symbols (i.e.,~they will immediately |
| 3067 |
% follow |\left| or |\right| in math mode). |
| 3068 |
% \begin{macrocode} |
| 3069 |
\let\bf@leftcurly=\{ |
| 3070 |
\let\bf@rightcurly=\} |
| 3071 |
% \end{macrocode} |
| 3072 |
% \end{macro} |
| 3073 |
% \end{macro} |
| 3074 |
% |
| 3075 |
% \begin{macro}{\bf@leftcurlystyle} |
| 3076 |
% \begin{macro}{\bf@rightcurlystyle} |
| 3077 |
% Define the default formatting for left and right curly braces as ``do |
| 3078 |
% nothing special''. |
| 3079 |
% \begin{macrocode} |
| 3080 |
\let\bf@leftcurlystyle=\relax |
| 3081 |
\let\bf@rightcurlystyle=\relax |
| 3082 |
% \end{macrocode} |
| 3083 |
% \end{macro} |
| 3084 |
% \end{macro} |
| 3085 |
% |
| 3086 |
% \begin{macro}{\curly@box} |
| 3087 |
% Define a box in which to temporarily store formatted curly braces. |
| 3088 |
% \changes{v1.2}{2004/06/14}{Bug fix: Defined \cs{curly@box} globally |
| 3089 |
% (suggested by Stefan Ulrich)} |
| 3090 |
% \begin{macrocode} |
| 3091 |
\newbox{\curly@box} |
| 3092 |
% \end{macrocode} |
| 3093 |
% \end{macro} |
| 3094 |
% |
| 3095 |
% \begin{macro}{\store@rcurly} |
| 3096 |
% \changes{v2.7}{2021/08/17}{Properly align right curly braces under |
| 3097 |
% Lua\string\LaTeX\ (bug reported by Georgi Nikiforov)} |
| 3098 |
% \begin{macro}{\curly@height} |
| 3099 |
% \begin{macro}{\half@curly@height} |
| 3100 |
% \begin{macro}{\curly@shift} |
| 3101 |
% \begin{macro}{\old@axis} |
| 3102 |
% Store a ``\}'' that's |#2| tall in box |#1|. |
| 3103 |
% The only unintuitive thing here is that we have to redefine |
| 3104 |
% |\fontdimen22|---axis height---to~0\,pt.\ before typesetting the curly |
| 3105 |
% brace. Otherwise, the brace would be vertically off-center by a few |
| 3106 |
% points. When we're finished, we reset it back to its old value. |
| 3107 |
% \begin{macrocode} |
| 3108 |
\def\store@rcurly#1#2{% |
| 3109 |
\begingroup |
| 3110 |
\bf@newdimen\curly@height |
| 3111 |
\setlength{\curly@height}{#2 - \bf@rightcurlyshrinkage}% |
| 3112 |
\bf@newdimen\half@curly@height |
| 3113 |
\setlength{\half@curly@height}{0.5\curly@height}% |
| 3114 |
\bf@newdimen\curly@shift |
| 3115 |
\setlength{\curly@shift}{\bf@rightcurlyshrinkage}% |
| 3116 |
\setlength{\curly@shift}{\half@curly@height + 0.5\curly@shift}% |
| 3117 |
\addtolength{\curly@shift}{-\fontdimen22\textfont2}% |
| 3118 |
\global\sbox{#1}{\raisebox{\curly@shift}{% |
| 3119 |
\bf@rightcurlystyle{% |
| 3120 |
$\left. |
| 3121 |
\vrule height\half@curly@height |
| 3122 |
width 0pt |
| 3123 |
depth\half@curly@height\right\bf@rightcurly$% |
| 3124 |
}% |
| 3125 |
}}% |
| 3126 |
\endgroup |
| 3127 |
} |
| 3128 |
% \end{macrocode} |
| 3129 |
% \end{macro} |
| 3130 |
% \end{macro} |
| 3131 |
% \end{macro} |
| 3132 |
% \end{macro} |
| 3133 |
% \end{macro} |
| 3134 |
% |
| 3135 |
% \begin{macro}{\store@lcurly} |
| 3136 |
% \changes{v2.7}{2021/08/17}{Properly align left curly braces under |
| 3137 |
% Lua\string\LaTeX\ (bug reported by Georgi Nikiforov)} |
| 3138 |
% \begin{macro}{\curly@height} |
| 3139 |
% \begin{macro}{\half@curly@height} |
| 3140 |
% \begin{macro}{\curly@shift} |
| 3141 |
% These are the same as |\store@rcurly|, etc.\ but using a ``\{'' |
| 3142 |
% instead of a ``\}''. |
| 3143 |
% \begin{macrocode} |
| 3144 |
\def\store@lcurly#1#2{% |
| 3145 |
\begingroup |
| 3146 |
\bf@newdimen\curly@height |
| 3147 |
\setlength{\curly@height}{#2 - \bf@leftcurlyshrinkage}% |
| 3148 |
\bf@newdimen\half@curly@height |
| 3149 |
\setlength{\half@curly@height}{0.5\curly@height}% |
| 3150 |
\bf@newdimen\curly@shift |
| 3151 |
\setlength{\curly@shift}{\bf@leftcurlyshrinkage}% |
| 3152 |
\setlength{\curly@shift}{\half@curly@height + 0.5\curly@shift}% |
| 3153 |
\addtolength{\curly@shift}{-\fontdimen22\textfont2}% |
| 3154 |
\global\sbox{#1}{\raisebox{\curly@shift}{% |
| 3155 |
\bf@leftcurlystyle{% |
| 3156 |
$\left\bf@leftcurly |
| 3157 |
\vrule height\half@curly@height |
| 3158 |
width 0pt |
| 3159 |
depth\half@curly@height\right.$% |
| 3160 |
}% |
| 3161 |
}}% |
| 3162 |
\endgroup |
| 3163 |
} |
| 3164 |
% \end{macrocode} |
| 3165 |
% \end{macro} |
| 3166 |
% \end{macro} |
| 3167 |
% \end{macro} |
| 3168 |
% \end{macro} |
| 3169 |
% |
| 3170 |
% \subsubsection{Right-side labels} |
| 3171 |
% |
| 3172 |
% \begin{macro}{\show@wordlabelr} |
| 3173 |
% This macro is output in the third column of every row of the |
| 3174 |
% |\ialign|ed bytefield table. It's normally a no-op, but |
| 3175 |
% |\end{rightwordgroup}| defines it to output the word label and then |
| 3176 |
% reset itself to a no-op. |
| 3177 |
% \begin{macrocode} |
| 3178 |
\def\show@wordlabelr{} |
| 3179 |
% \end{macrocode} |
| 3180 |
% \end{macro} |
| 3181 |
% |
| 3182 |
% \begin{macro}{\wordlabelr@start} |
| 3183 |
% \begin{macro}{\wordlabelr@end} |
| 3184 |
% Declare the starting and ending height (in points) of the set of rows |
| 3185 |
% to be labeled on the right. |
| 3186 |
% \begin{macrocode} |
| 3187 |
\newlength{\wordlabelr@start} |
| 3188 |
\newlength{\wordlabelr@end} |
| 3189 |
% \end{macrocode} |
| 3190 |
% \end{macro} |
| 3191 |
% \end{macro} |
| 3192 |
% |
| 3193 |
% \begin{environment}{rightwordgroup} |
| 3194 |
% \usermacro |
| 3195 |
% Label the words defined between |\begin{rightwordgroup}| and |
| 3196 |
% |\end{rightwordgroup}| on the right side of the bit field. The first, |
| 3197 |
% optional, argument is a list of parameters, as defined in |
| 3198 |
% Section~\ref{sec:options}. The second, mandatory, argument is the |
| 3199 |
% text of the label. The label is typeset to the right of a large curly |
| 3200 |
% brace, which groups the words together. |
| 3201 |
% \changes{v2.6}{2020/10/29}{Suppress spaces following the |
| 3202 |
% \protect\cs{end}\protect\texttt{\{rightwordgroup\}}} |
| 3203 |
% \changes{v2.6}{2020/10/31}{Accept key/value options} |
| 3204 |
% \begin{macrocode} |
| 3205 |
\newenvironment{rightwordgroup}[2][]{% |
| 3206 |
% \end{macrocode} |
| 3207 |
% We begin by ending the group that |\begin{rightwordgroup}| created. This |
| 3208 |
% lets the |rightwordgroup| environment span rows (because we're technically |
| 3209 |
% no longer within the environment). |
| 3210 |
% \begin{macrocode} |
| 3211 |
\endgroup |
| 3212 |
% \end{macrocode} |
| 3213 |
% \begin{macro}{\wordlabelr@start} |
| 3214 |
% \begin{macro}{\wordlabelr@params} |
| 3215 |
% \begin{macro}{\wordlabelr@text} |
| 3216 |
% |\begin{rightwordgroup}| merely stores the starting height in |
| 3217 |
% |\wordlabelr@start| and the user-supplied text in |\wordlabelr@text|. |
| 3218 |
% |\end{rightwordgroup}| does most of the work. |
| 3219 |
% \begin{macrocode} |
| 3220 |
\global\wordlabelr@start=\bytefield@height |
| 3221 |
\gdef\wordlabelr@params{#1}% |
| 3222 |
\gdef\wordlabelr@text{#2}% |
| 3223 |
\ignorespaces |
| 3224 |
}{% |
| 3225 |
% \end{macrocode} |
| 3226 |
% \end{macro} |
| 3227 |
% \end{macro} |
| 3228 |
% \end{macro} |
| 3229 |
% \begin{macro}{\wordlabelr@end} |
| 3230 |
% Because we already ended the group that |\begin{rightwordgroup}| |
| 3231 |
% created we now have to begin a group for |\end{rightwordgroup}| to |
| 3232 |
% end. |
| 3233 |
% \begin{macrocode} |
| 3234 |
\begingroup |
| 3235 |
\global\wordlabelr@end=\bytefield@height |
| 3236 |
% \end{macrocode} |
| 3237 |
% \end{macro} |
| 3238 |
% \begin{macro}{\show@wordlabelr} |
| 3239 |
% Redefine |\show@wordlabelr| to output |\bf@rightcurlyspace| space, |
| 3240 |
% followed by a large curly brace (in |\curlybox|), followed by |
| 3241 |
% |\bf@rightlabelspace| space, followed by the user's text (previously |
| 3242 |
% recorded in |\wordlabelr@text|). We typeset |\wordlabelr@text| within |
| 3243 |
% a |tabular| environment, so \LaTeX\ will calculate its width |
| 3244 |
% automatically. |
| 3245 |
% \begin{macrocode} |
| 3246 |
\gdef\show@wordlabelr{% |
| 3247 |
\sbox{\word@label@box}{% |
| 3248 |
\begin{tabular}[b]{@{}l@{}}\wordlabelr@text\end{tabular}% |
| 3249 |
}% |
| 3250 |
\settowidth{\label@box@width}{\usebox{\word@label@box}}% |
| 3251 |
\setlength{\label@box@height}{\wordlabelr@end-\wordlabelr@start}% |
| 3252 |
% \end{macrocode} |
| 3253 |
% Evaluate any parameters passed to |\begin{rightwordgroup}| right |
| 3254 |
% before we render the curly brace. |
| 3255 |
% \begin{macrocode} |
| 3256 |
\expandafter\bf@bytefieldsetup\expandafter{\wordlabelr@params}% |
| 3257 |
\store@rcurly{\curly@box}{\label@box@height}% |
| 3258 |
\bf@newdimen\total@box@width |
| 3259 |
\setlength{\total@box@width}{% |
| 3260 |
\bf@rightcurlyspace + |
| 3261 |
\widthof{\usebox{\curly@box}} + |
| 3262 |
\bf@rightlabelspace + |
| 3263 |
\label@box@width |
| 3264 |
}% |
| 3265 |
\begin{picture}(\strip@pt\total@box@width,0) |
| 3266 |
\put(0,0){% |
| 3267 |
\hspace*{\bf@rightcurlyspace}% |
| 3268 |
\usebox{\curly@box}% |
| 3269 |
\hspace*{\bf@rightlabelspace}% |
| 3270 |
\makebox(\strip@pt\label@box@width,\strip@pt\label@box@height){% |
| 3271 |
\usebox{\word@label@box}% |
| 3272 |
}% |
| 3273 |
}% |
| 3274 |
\end{picture}% |
| 3275 |
% \end{macrocode} |
| 3276 |
% The last thing |\show@wordlabelr| does is redefine itself back to a no-op. |
| 3277 |
% \begin{macrocode} |
| 3278 |
\gdef\show@wordlabelr{}}% |
| 3279 |
% \end{macrocode} |
| 3280 |
% \end{macro} |
| 3281 |
% \begin{macro}{\@currenvir} |
| 3282 |
% Because of our meddling with |\begingroup| and |\endgroup|, the |
| 3283 |
% current environment is all messed up. We therefore force the |
| 3284 |
% |\end{rightwordgroup}| to succeed, even if it doesn't match the preceding |
| 3285 |
% |\begin|. |
| 3286 |
% \begin{macrocode} |
| 3287 |
\def\@currenvir{rightwordgroup}% |
| 3288 |
\ignorespacesafterend |
| 3289 |
} |
| 3290 |
% \end{macrocode} |
| 3291 |
% \end{macro} |
| 3292 |
% \end{environment} |
| 3293 |
% |
| 3294 |
% \subsubsection{Left-side labels} |
| 3295 |
% |
| 3296 |
% \begin{macro}{\wordlabell@start} |
| 3297 |
% \begin{macro}{\wordlabell@end} |
| 3298 |
% Declare the starting and ending height (in points) of the set of rows |
| 3299 |
% to be labeled on the left. |
| 3300 |
% \begin{macrocode} |
| 3301 |
\newlength{\wordlabell@start} |
| 3302 |
\newlength{\wordlabell@end} |
| 3303 |
% \end{macrocode} |
| 3304 |
% \end{macro} |
| 3305 |
% \end{macro} |
| 3306 |
% |
| 3307 |
% \begin{macro}{\total@box@width} |
| 3308 |
% Declare the total width of the next label to typeset on the left of |
| 3309 |
% the bit field, that is, the aggregate width of the text box, curly |
| 3310 |
% brace, and spaces on either side of the curly brace. |
| 3311 |
% \begin{macrocode} |
| 3312 |
\newlength{\total@lbox@width} |
| 3313 |
% \end{macrocode} |
| 3314 |
% \end{macro} |
| 3315 |
% |
| 3316 |
% \begin{macro}{\make@lspace} |
| 3317 |
% This macro is output in the first column of every row of the |
| 3318 |
% |\ialign|ed bytefield table. It's normally a no-op, but |
| 3319 |
% |\begin{leftwordgroup}| defines it to output enough space for the next word |
| 3320 |
% label and then reset itself to a no-op. |
| 3321 |
% \begin{macrocode} |
| 3322 |
\gdef\make@lspace{} |
| 3323 |
% \end{macrocode} |
| 3324 |
% \end{macro} |
| 3325 |
% |
| 3326 |
% \begin{environment}{leftwordgroup} |
| 3327 |
% \usermacro |
| 3328 |
% This environment is essentially the same as the |rightwordgroup| |
| 3329 |
% environment but puts the label on the left. However, the following |
| 3330 |
% code is not symmetric to that of |rightwordgroup|. The problem is that we |
| 3331 |
% encounter |\begin{leftwordgroup}| after entering the second |
| 3332 |
% (i.e.,~figure) column, which doesn't give us a chance to reserve space |
| 3333 |
% in the first (i.e.,~left label) column. When we reach the |
| 3334 |
% |\end{leftwordgroup}|, we know the height of the group of words we wish |
| 3335 |
% to label. However, if we try to label the words in the subsequent |
| 3336 |
% first column, we won't know the vertical offset from the ``cursor'' at |
| 3337 |
% which to start drawing the label, because we can't know the height of |
| 3338 |
% the subsequent row until we reach the second |
| 3339 |
% column.\footnote{Question: Is there a way to push the label up to the |
| 3340 |
% \emph{top} of the subsequent row, perhaps with |
| 3341 |
% \texttt{\string\vfill}?} |
| 3342 |
% |
| 3343 |
% Our solution is to allocate space for the box the next time we enter a |
| 3344 |
% first column. As long as space is eventually allocated, the column |
| 3345 |
% will expand to fit that space. |\end{leftwordgroup}| outputs the label |
| 3346 |
% immediately. Even though |\end{leftwordgroup}| is called at the end of |
| 3347 |
% the \emph{second} column, it |\put|s the label at a sufficiently |
| 3348 |
% negative $x$ location for it to overlap the first column. Because |
| 3349 |
% there will eventually be enough space to accomodate the label, we know |
| 3350 |
% that the label won't overlap the bit field or extend beyond the bit-field |
| 3351 |
% boundaries. |
| 3352 |
% \changes{v2.6}{2020/10/29}{Suppress spaces following the |
| 3353 |
% \protect\cs{end}\protect\texttt{\{leftwordgroup\}}} |
| 3354 |
% \changes{v2.6}{2020/10/31}{Accept key/value options} |
| 3355 |
% \begin{macrocode} |
| 3356 |
\newenvironment{leftwordgroup}[2][]{% |
| 3357 |
% \end{macrocode} |
| 3358 |
% \begin{macro}{\wordlabell@start} |
| 3359 |
% \begin{macro}{\wordlabell@params} |
| 3360 |
% \begin{macro}{\wordlabell@text} |
| 3361 |
% We store the starting height, optional parameters (see |
| 3362 |
% Section~\ref{sec:options}), and label text, all of which are needed by |
| 3363 |
% the |\end{leftwordgroup}|. We immediately parse the parameters |
| 3364 |
% because they may affect the \cs{store@lcurly} invocation below. |
| 3365 |
% \begin{macrocode} |
| 3366 |
\global\wordlabell@start=\bytefield@height |
| 3367 |
\gdef\wordlabell@params{#1}% |
| 3368 |
\gdef\wordlabell@text{#2}% |
| 3369 |
\bf@bytefieldsetup{#1}% |
| 3370 |
% \end{macrocode} |
| 3371 |
% \end{macro} |
| 3372 |
% \end{macro} |
| 3373 |
% \end{macro} |
| 3374 |
% Next, we typeset a draft version of the label into |\word@label@box|, |
| 3375 |
% which we measure (into |\total@lbox@width|) and then discard. |
| 3376 |
% We can't typeset the final version of the label until we reach the |
| 3377 |
% |\end{leftwordgroup}|, because that's when we learn the height of the |
| 3378 |
% word group. Without knowing the height of the word group, we don't |
| 3379 |
% how how big to make the curly brace. In the scratch version, we make |
| 3380 |
% the curly brace 5\,cm.~tall. This should be more than large enough to |
| 3381 |
% reach the maximum curly-brace width, which is all we really care about |
| 3382 |
% at this point. |
| 3383 |
% \begin{macrocode} |
| 3384 |
\sbox{\word@label@box}{% |
| 3385 |
\begin{tabular}[b]{@{}l@{}}\wordlabell@text\end{tabular}}% |
| 3386 |
\settowidth{\label@box@width}{\usebox{\word@label@box}}% |
| 3387 |
\store@lcurly{\curly@box}{5cm}% |
| 3388 |
\setlength{\total@lbox@width}{% |
| 3389 |
\bf@leftcurlyspace + |
| 3390 |
\widthof{\usebox{\curly@box}} + |
| 3391 |
\bf@leftlabelspace + |
| 3392 |
\label@box@width}% |
| 3393 |
\global\total@lbox@width=\total@lbox@width |
| 3394 |
% \end{macrocode} |
| 3395 |
% \begin{macro}{\make@lspace} |
| 3396 |
% Now we know how wide the box is going to be (unless, of course, |
| 3397 |
% the user is using some weird math font that scales the width of a curly |
| 3398 |
% brace proportionally to its height). So we redefine |\make@lspace| to |
| 3399 |
% output |\total@lbox@width|'s worth of space and then redefine itself |
| 3400 |
% back to a no-op. |
| 3401 |
% \begin{macrocode} |
| 3402 |
\gdef\make@lspace{% |
| 3403 |
\hspace*{\total@lbox@width}% |
| 3404 |
\gdef\make@lspace{}% |
| 3405 |
}% |
| 3406 |
% \end{macrocode} |
| 3407 |
% We now end the group that |\begin{rightwordgroup}| created. This lets |
| 3408 |
% the |leftwordgroup| environment span rows (because we're technically |
| 3409 |
% no longer within the environment). |
| 3410 |
% \begin{macrocode} |
| 3411 |
\endgroup |
| 3412 |
\ignorespaces |
| 3413 |
}{% |
| 3414 |
% \end{macrocode} |
| 3415 |
% \end{macro} |
| 3416 |
% Because we already ended the group that |\begin{leftwordgroup}| created |
| 3417 |
% we have to start the |\end{leftwordgroup}| by beginning a group for |
| 3418 |
% |\end{leftwordgroup}| to end. |
| 3419 |
% \begin{macrocode} |
| 3420 |
\begingroup |
| 3421 |
% \end{macrocode} |
| 3422 |
% The |\end{leftwordgroup}| code is comparatively straightforward. We |
| 3423 |
% calculate the final height of the word group, and then output the |
| 3424 |
% label text, followed by |\bf@leftlabelspace| space, followed by a |
| 3425 |
% curly brace (now that we know how tall it's supposed to be), followed |
| 3426 |
% by |\bf@leftcurlyspace| space. The trick, as described earlier, is |
| 3427 |
% that we typeset the entire label in the second column, but in a $0 |
| 3428 |
% \times 0$ |picture| environment and with a negative horizontal offset |
| 3429 |
% (|\starting@point|), thereby making it overlap the first column. |
| 3430 |
% Before typesetting the curly brace we re-parse the optional parameters |
| 3431 |
% because we're in a new group from the one in which we parsed them |
| 3432 |
% before, and the parameters can affect the second \cs{store@lcurly} |
| 3433 |
% invocation just they could have affected the first. |
| 3434 |
% \begin{macrocode} |
| 3435 |
\global\wordlabell@end=\bytefield@height |
| 3436 |
\bf@newdimen\starting@point |
| 3437 |
\setlength{\starting@point}{% |
| 3438 |
-\total@lbox@width - \bf@bitwidth*\bits@wide}% |
| 3439 |
\sbox{\word@label@box}{% |
| 3440 |
\begin{tabular}[b]{@{}l@{}}\wordlabell@text\end{tabular}}% |
| 3441 |
\settowidth{\label@box@width}{\usebox{\word@label@box}}% |
| 3442 |
\setlength{\label@box@height}{\wordlabell@end-\wordlabell@start}% |
| 3443 |
\expandafter\bf@bytefieldsetup\expandafter{\wordlabell@params}% |
| 3444 |
\store@lcurly{\curly@box}{\label@box@height}% |
| 3445 |
\begin{picture}(0,0) |
| 3446 |
\put(\strip@pt\starting@point,0){% |
| 3447 |
\makebox(\strip@pt\label@box@width,\strip@pt\label@box@height){% |
| 3448 |
\usebox{\word@label@box}}% |
| 3449 |
\hspace*{\bf@leftlabelspace}% |
| 3450 |
\usebox{\curly@box}% |
| 3451 |
\hspace*{\bf@leftcurlyspace}} |
| 3452 |
\end{picture}% |
| 3453 |
% \end{macrocode} |
| 3454 |
% \begin{macro}{\@currenvir} |
| 3455 |
% Because of our meddling with |\begingroup| and |\endgroup|, the |
| 3456 |
% current environment is all messed up. We therefore force the |
| 3457 |
% |\end{leftwordgroup}| to succeed, even if it doesn't match the preceding |
| 3458 |
% |\begin|. |
| 3459 |
% \begin{macrocode} |
| 3460 |
\def\@currenvir{leftwordgroup}% |
| 3461 |
\ignorespacesafterend |
| 3462 |
} |
| 3463 |
% \end{macrocode} |
| 3464 |
% \end{macro} |
| 3465 |
% \end{environment} |
| 3466 |
% |
| 3467 |
% \subsubsection{Scratch space} |
| 3468 |
% |
| 3469 |
% \begin{macro}{\label@box@width} |
| 3470 |
% \begin{macro}{\label@box@height} |
| 3471 |
% \begin{macro}{\word@label@box} |
| 3472 |
% Declare some scratch storage for the width, height, and contents of |
| 3473 |
% the word label we're about to output. |
| 3474 |
% \begin{macrocode} |
| 3475 |
\newlength{\label@box@width} |
| 3476 |
\newlength{\label@box@height} |
| 3477 |
\newsavebox{\word@label@box} |
| 3478 |
% \end{macrocode} |
| 3479 |
% \end{macro} |
| 3480 |
% \end{macro} |
| 3481 |
% \end{macro} |
| 3482 |
% |
| 3483 |
% \subsection{Compatibility mode} |
| 3484 |
% |
| 3485 |
% \begin{macro}{\bf@enter@compatibility@mode@i} |
| 3486 |
% \pkgname{bytefield}'s interface changed substantially with the move to |
| 3487 |
% version~2.0. To give version~1.\textit{x} users a quick way to build |
| 3488 |
% their old documents, we provide a version~1.\textit{x} compatibility |
| 3489 |
% mode. We don't enable this by default because it exposes a number of |
| 3490 |
% extra length registers (a precious resource) and because we want to |
| 3491 |
% encourage users to migrate to the new interface. |
| 3492 |
% \begin{macrocode} |
| 3493 |
\newcommand{\bf@enter@compatibility@mode@i}{% |
| 3494 |
% \end{macrocode} |
| 3495 |
% \begin{macro}{\bitwidth} |
| 3496 |
% \begin{macro}{\byteheight} |
| 3497 |
% \begin{macro}{\curlyspace} |
| 3498 |
% \begin{macro}{\labelspace} |
| 3499 |
% \begin{macro}{\curlyshrinkage} |
| 3500 |
% Define a handful of lengths that the user was allowed to |\setlength| |
| 3501 |
% explicitly in \pkgname{bytefield}~1.\textit{x}. |
| 3502 |
% \begin{macrocode} |
| 3503 |
\PackageInfo{bytefield}{Entering version 1 compatibility mode}% |
| 3504 |
\newlength{\bitwidth}% |
| 3505 |
\newlength{\byteheight}% |
| 3506 |
\newlength{\curlyspace}% |
| 3507 |
\newlength{\labelspace}% |
| 3508 |
\newlength{\curlyshrinkage}% |
| 3509 |
% \end{macrocode} |
| 3510 |
% \begin{macrocode} |
| 3511 |
\settowidth{\bitwidth}{\tiny 99i}% |
| 3512 |
\setlength{\byteheight}{4ex}% |
| 3513 |
\setlength{\curlyspace}{1ex}% |
| 3514 |
\setlength{\labelspace}{0.5ex}% |
| 3515 |
\setlength{\curlyshrinkage}{5pt}% |
| 3516 |
% \end{macrocode} |
| 3517 |
% \end{macro} |
| 3518 |
% \end{macro} |
| 3519 |
% \end{macro} |
| 3520 |
% \end{macro} |
| 3521 |
% \end{macro} |
| 3522 |
% \begin{macro}{\newbytefield} |
| 3523 |
% \begin{macro}{\endnewbytefield} |
| 3524 |
% \begin{environment}{bytefield} |
| 3525 |
% Redefine the |bytefield| environment in terms of the existing |
| 3526 |
% (new-interface) |bytefield| environment. The difference is that the |
| 3527 |
% redefinition utilizes all of the preceding lengths. |
| 3528 |
% \begin{macrocode} |
| 3529 |
\let\newbytefield=\bytefield |
| 3530 |
\let\endnewbytefield=\endbytefield |
| 3531 |
\renewenvironment{bytefield}[1]{% |
| 3532 |
\begin{newbytefield}[% |
| 3533 |
bitwidth=\bitwidth, |
| 3534 |
bitheight=\byteheight, |
| 3535 |
curlyspace=\curlyspace, |
| 3536 |
labelspace=\labelspace, |
| 3537 |
curlyshrinkage=\curlyshrinkage]{##1}% |
| 3538 |
}{% |
| 3539 |
\end{newbytefield}% |
| 3540 |
} |
| 3541 |
% \end{macrocode} |
| 3542 |
% \end{environment} |
| 3543 |
% \end{macro} |
| 3544 |
% \end{macro} |
| 3545 |
% \begin{macro}{\wordgroupr} |
| 3546 |
% \begin{macro}{\endwordgroupr} |
| 3547 |
% \begin{macro}{\wordgroupl} |
| 3548 |
% \begin{macro}{\endwordgroupl} |
| 3549 |
% Define |\wordgroupr|, |\endwordgroupr|, |\wordgroupl|, and |
| 3550 |
% |\endwordgroupl| in terms of the new |rightwordgroup| and |
| 3551 |
% |leftwordgroup| environments. |
| 3552 |
% \begin{macrocode} |
| 3553 |
\def\wordgroupr{\begin{rightwordgroup}} |
| 3554 |
\def\endwordgroupr{\end{rightwordgroup}} |
| 3555 |
\def\wordgroupl{\begin{leftwordgroup}} |
| 3556 |
\def\endwordgroupl{\end{leftwordgroup}} |
| 3557 |
% \end{macrocode} |
| 3558 |
% \end{macro} |
| 3559 |
% \end{macro} |
| 3560 |
% \end{macro} |
| 3561 |
% \end{macro} |
| 3562 |
% |
| 3563 |
% \begin{macro}{\bytefieldsetup} |
| 3564 |
% Disable |\bytefieldsetup| in compatibility mode because it doesn't |
| 3565 |
% work as expected. (Every use of the compatibility-mode |bytefield| |
| 3566 |
% environment overwrites all of the figure-formatting values.) |
| 3567 |
% \begin{macrocode} |
| 3568 |
\renewcommand{\bytefieldsetup}[1]{% |
| 3569 |
\PackageError{bytefield}{% |
| 3570 |
The \protect\bytefieldsetup\space macro is not available in\MessageBreak |
| 3571 |
version 1 compatibility mode% |
| 3572 |
}{% |
| 3573 |
Remove [compat1] from the \protect\usepackage{bytefield} line to |
| 3574 |
make \protect\bytefieldsetup\MessageBreak |
| 3575 |
available to this document.\space\space (The document may also need |
| 3576 |
to be modified to use\MessageBreak |
| 3577 |
the new bytefield interface.) |
| 3578 |
}% |
| 3579 |
}% |
| 3580 |
} |
| 3581 |
% \end{macrocode} |
| 3582 |
% \end{macro} |
| 3583 |
% \end{macro} |
| 3584 |
% |
| 3585 |
% \begin{macro}{\wordgroupr} |
| 3586 |
% \begin{macro}{\endwordgroupr} |
| 3587 |
% \begin{macro}{\wordgroupl} |
| 3588 |
% \begin{macro}{\endwordgroupl} |
| 3589 |
% Issue a helpful error message for the commands that were removed in |
| 3590 |
% \pkgname{bytefield}~v2.0. While this won't help users whose first |
| 3591 |
% invalid action is to modify a no-longer-extant length register such as |
| 3592 |
% |\bitwidth| or |\byteheight|, it may benefit at least a few users who |
| 3593 |
% didn't realize that the \pkgname{bytefield} interface has changed |
| 3594 |
% substantially with version~2.0. |
| 3595 |
% \begin{macrocode} |
| 3596 |
\newcommand{\wordgroupr}{% |
| 3597 |
\PackageError{bytefield}{% |
| 3598 |
Macros \protect\wordgroupr, \protect\wordgroupl, \protect\endwordgroupr, |
| 3599 |
\MessageBreak |
| 3600 |
and \protect\endwordgroupl\space no longer exist% |
| 3601 |
}{% |
| 3602 |
Starting with version 2.0, bytefield uses \protect\begin{wordgroupr}... |
| 3603 |
\MessageBreak |
| 3604 |
\protect\end{wordgroupr} and \protect\begin{wordgroupl}...% |
| 3605 |
\protect\end{wordgroupl}\MessageBreak |
| 3606 |
to specify word groups and a new \protect\bytefieldsetup\space macro to |
| 3607 |
\MessageBreak |
| 3608 |
change bytefield's various formatting parameters.% |
| 3609 |
}% |
| 3610 |
} |
| 3611 |
\let\endwordgroupr=\wordgroupr |
| 3612 |
\let\wordgroupl=\wordgroupr |
| 3613 |
\let\endwordgroupl=\wordgroupr |
| 3614 |
% \end{macrocode} |
| 3615 |
% \end{macro} |
| 3616 |
% \end{macro} |
| 3617 |
% \end{macro} |
| 3618 |
% \end{macro} |
| 3619 |
% |
| 3620 |
% |
| 3621 |
% \subsection{Option processing} |
| 3622 |
% |
| 3623 |
% We use the \textsf{keyval} package to handle option processing. |
| 3624 |
% Because all of \pkgname{bytefield}'s options have local impact, options |
| 3625 |
% can be specified either as package arguments or through the use of the |
| 3626 |
% |\bytefieldsetup| macro. |
| 3627 |
% |
| 3628 |
% \begin{macro}{\KV@bytefield@bitwidth} |
| 3629 |
% \begin{macro}{\bf@bw@arg} |
| 3630 |
% \begin{macro}{\bf@auto} |
| 3631 |
% Specify the width of a bit number in the bit header. If the special |
| 3632 |
% value ``|auto|'' is given, set the width to the width of a formatted |
| 3633 |
% ``|99i|''. |
| 3634 |
% \begin{macrocode} |
| 3635 |
\define@key{bytefield}{bitwidth}{% |
| 3636 |
\def\bf@bw@arg{#1}% |
| 3637 |
\def\bf@auto{auto}% |
| 3638 |
\ifx\bf@bw@arg\bf@auto |
| 3639 |
\settowidth{\bf@bitwidth}{\bf@bitformatting{99i}}% |
| 3640 |
\else |
| 3641 |
\setlength{\bf@bitwidth}{#1}% |
| 3642 |
\fi |
| 3643 |
} |
| 3644 |
% \end{macrocode} |
| 3645 |
% \end{macro} |
| 3646 |
% \end{macro} |
| 3647 |
% \end{macro} |
| 3648 |
% |
| 3649 |
% \begin{macro}{\KV@bytefield@bf@bitheight} |
| 3650 |
% Specify the height of a bit in a |\bitbox| or |\wordbox|. |
| 3651 |
% \begin{macrocode} |
| 3652 |
\define@key{bytefield}{bitheight}{\setlength{\bf@bitheight}{#1}} |
| 3653 |
% \end{macrocode} |
| 3654 |
% \end{macro} |
| 3655 |
% |
| 3656 |
% \begin{macro}{\KV@bytefield@bitformatting} |
| 3657 |
% Specify the style of a bit number in the bit header. This should be |
| 3658 |
% passed an expression that takes either one argument (e.g.,~|\textit|) |
| 3659 |
% or no arguments (e.g.,~|{\small\bfseries}|). |
| 3660 |
% \begin{macrocode} |
| 3661 |
\define@key{bytefield}{bitformatting}{\def\bf@bitformatting{#1}} |
| 3662 |
% \end{macrocode} |
| 3663 |
% \end{macro} |
| 3664 |
% |
| 3665 |
% \begin{macro}{\KV@bytefield@boxformatting} |
| 3666 |
% Specify a style to be applied to the contents of every bit box and |
| 3667 |
% word box. This should be passed an expression that takes either one |
| 3668 |
% argument (e.g.,~|\textit|) or no arguments |
| 3669 |
% (e.g.,~|{\small\bfseries}|). |
| 3670 |
% \begin{macrocode} |
| 3671 |
\define@key{bytefield}{boxformatting}{\def\bf@boxformatting{#1}} |
| 3672 |
% \end{macrocode} |
| 3673 |
% \end{macro} |
| 3674 |
% |
| 3675 |
% \begin{macro}{\KV@bytefield@leftcurly} |
| 3676 |
% \begin{macro}{\KV@bytefield@rightcurly} |
| 3677 |
% \begin{macro}{\bf@leftcurly} |
| 3678 |
% \begin{macro}{\bf@rightcurly} |
| 3679 |
% Specify the symbol to use for bracketing a left or right word group. |
| 3680 |
% This must be an extensible math delimiter (i.e.,~something that can |
| 3681 |
% immediately follow |\left| or |\right| in math mode). |
| 3682 |
% \begin{macrocode} |
| 3683 |
\define@key{bytefield}{leftcurly}{\def\bf@leftcurly{#1}} |
| 3684 |
\define@key{bytefield}{rightcurly}{\def\bf@rightcurly{#1}} |
| 3685 |
% \end{macrocode} |
| 3686 |
% \end{macro} |
| 3687 |
% \end{macro} |
| 3688 |
% \end{macro} |
| 3689 |
% \end{macro} |
| 3690 |
% |
| 3691 |
% \begin{macro}{\KV@bytefield@leftcurlyspace} |
| 3692 |
% \begin{macro}{\KV@bytefield@rightcurlyspace} |
| 3693 |
% \begin{macro}{\KV@bytefield@curlyspace} |
| 3694 |
% \begin{macro}{\bf@leftcurlyspace} |
| 3695 |
% \begin{macro}{\bf@rightcurlyspace} |
| 3696 |
% Specify the amount of space between the bit fields in a word group and |
| 3697 |
% the adjacent left or right curly brace. The \optname{curlyspace} |
| 3698 |
% option is a shortcut that puts the same space before both left and |
| 3699 |
% right curly braces. |
| 3700 |
% \begin{macrocode} |
| 3701 |
\define@key{bytefield}{leftcurlyspace}{\def\bf@leftcurlyspace{#1}} |
| 3702 |
\define@key{bytefield}{rightcurlyspace}{\def\bf@rightcurlyspace{#1}} |
| 3703 |
\define@key{bytefield}{curlyspace}{% |
| 3704 |
\def\bf@leftcurlyspace{#1}% |
| 3705 |
\def\bf@rightcurlyspace{#1}% |
| 3706 |
} |
| 3707 |
% \end{macrocode} |
| 3708 |
% \end{macro} |
| 3709 |
% \end{macro} |
| 3710 |
% \end{macro} |
| 3711 |
% \end{macro} |
| 3712 |
% \end{macro} |
| 3713 |
% |
| 3714 |
% \begin{macro}{\KV@bytefield@leftlabelspace} |
| 3715 |
% \begin{macro}{\KV@bytefield@rightlabelspace} |
| 3716 |
% \begin{macro}{\KV@bytefield@labelspace} |
| 3717 |
% \begin{macro}{\bf@leftlabelspace} |
| 3718 |
% \begin{macro}{\bf@rightlabelspace} |
| 3719 |
% Specify the amount of space between a left or right word group's curly |
| 3720 |
% brace and the associated label text. The |labelspace| option is a |
| 3721 |
% shortcut that puts the same space after both left and right curly |
| 3722 |
% braces. |
| 3723 |
% \begin{macrocode} |
| 3724 |
\define@key{bytefield}{leftlabelspace}{\def\bf@leftlabelspace{#1}} |
| 3725 |
\define@key{bytefield}{rightlabelspace}{\def\bf@rightlabelspace{#1}} |
| 3726 |
\define@key{bytefield}{labelspace}{% |
| 3727 |
\def\bf@leftlabelspace{#1}% |
| 3728 |
\def\bf@rightlabelspace{#1}% |
| 3729 |
} |
| 3730 |
% \end{macrocode} |
| 3731 |
% \end{macro} |
| 3732 |
% \end{macro} |
| 3733 |
% \end{macro} |
| 3734 |
% \end{macro} |
| 3735 |
% \end{macro} |
| 3736 |
% |
| 3737 |
% \begin{macro}{\KV@bytefield@leftcurlyshrinkage} |
| 3738 |
% \begin{macro}{\KV@bytefield@rightcurlyshrinkage} |
| 3739 |
% \begin{macro}{\KV@bytefield@curlyshrinkage} |
| 3740 |
% \begin{macro}{\bf@leftcurlyshrinkage} |
| 3741 |
% \begin{macro}{\bf@rightcurlyshrinkage} |
| 3742 |
% Specify the number of points by which to reduce the height of a curly |
| 3743 |
% brace (left, right, or both) so its ends don't overlap whatever's |
| 3744 |
% above or below it. |
| 3745 |
% \begin{macrocode} |
| 3746 |
\define@key{bytefield}{leftcurlyshrinkage}{\def\bf@leftcurlyshrinkage{#1}} |
| 3747 |
\define@key{bytefield}{rightcurlyshrinkage}{\def\bf@rightcurlyshrinkage{#1}} |
| 3748 |
\define@key{bytefield}{curlyshrinkage}{% |
| 3749 |
\def\bf@leftcurlyshrinkage{#1}% |
| 3750 |
\def\bf@rightcurlyshrinkage{#1}% |
| 3751 |
} |
| 3752 |
% \end{macrocode} |
| 3753 |
% \end{macro} |
| 3754 |
% \end{macro} |
| 3755 |
% \end{macro} |
| 3756 |
% \end{macro} |
| 3757 |
% \end{macro} |
| 3758 |
% |
| 3759 |
% \begin{macro}{\KV@bytefield@leftcurlystyle} |
| 3760 |
% \begin{macro}{\KV@bytefield@rightcurlystyle} |
| 3761 |
% \begin{macro}{\KV@bytefieldcurlystyle} |
| 3762 |
% \begin{macro}{\bf@leftcurlystyle} |
| 3763 |
% \begin{macro}{\bf@rightcurlystyle} |
| 3764 |
% Specify a macro that takes either zero or one argument and that |
| 3765 |
% precedes the text that draws a left curly brace, right curly brace, or |
| 3766 |
% either curly brace. |
| 3767 |
% \begin{macrocode} |
| 3768 |
\define@key{bytefield}{leftcurlystyle}{\def\bf@leftcurlystyle{#1}} |
| 3769 |
\define@key{bytefield}{rightcurlystyle}{\def\bf@rightcurlystyle{#1}} |
| 3770 |
\define@key{bytefield}{curlystyle}{% |
| 3771 |
\def\bf@leftcurlystyle{#1}% |
| 3772 |
\def\bf@rightcurlystyle{#1}% |
| 3773 |
} |
| 3774 |
% \end{macrocode} |
| 3775 |
% \end{macro} |
| 3776 |
% \end{macro} |
| 3777 |
% \end{macro} |
| 3778 |
% \end{macro} |
| 3779 |
% \end{macro} |
| 3780 |
% |
| 3781 |
% \begin{macro}{\KV@bytefield@endianness} |
| 3782 |
% \begin{macro}{\bf@parse@endianness} |
| 3783 |
% Set the default endianness to either little endian or big endian. |
| 3784 |
% \begin{macrocode} |
| 3785 |
\define@key{bytefield}{endianness}{\bf@parse@endianness{#1}} |
| 3786 |
% \end{macrocode} |
| 3787 |
% \begin{macrocode} |
| 3788 |
\newcommand{\bf@parse@endianness}[1]{% |
| 3789 |
\def\bf@little{little}% |
| 3790 |
\def\bf@big{big}% |
| 3791 |
\def\bf@arg{#1}% |
| 3792 |
\ifx\bf@arg\bf@little |
| 3793 |
\def\bf@bit@endianness{l}% |
| 3794 |
\else |
| 3795 |
\ifx\bf@arg\bf@big |
| 3796 |
\def\bf@bit@endianness{b}% |
| 3797 |
\else |
| 3798 |
\PackageError{bytefield}{% |
| 3799 |
Invalid argument "#1" to the endianness option% |
| 3800 |
}{% |
| 3801 |
The endianness option must be set to either "little" or |
| 3802 |
"big".\MessageBreak |
| 3803 |
Please specify either endianness=little or endianness=big. |
| 3804 |
}% |
| 3805 |
\fi |
| 3806 |
\fi |
| 3807 |
} |
| 3808 |
% \end{macrocode} |
| 3809 |
% \end{macro} |
| 3810 |
% \end{macro} |
| 3811 |
% |
| 3812 |
% \begin{macro}{\KV@bytefield@lsb} |
| 3813 |
% Specify a numerical value for the least significant bit of a word. |
| 3814 |
% \begin{macrocode} |
| 3815 |
\define@key{bytefield}{lsb}{\def\bf@first@bit{#1}} |
| 3816 |
% \end{macrocode} |
| 3817 |
% \end{macro} |
| 3818 |
% |
| 3819 |
% \begin{macro}{\bf@bgcolor} |
| 3820 |
% \begin{macro}{\KV@bytefield@bgcolor} |
| 3821 |
% Specify a background color for a bit box or word box. |
| 3822 |
% \begin{macrocode} |
| 3823 |
\define@key{bytefield}{bgcolor}{\def\bf@bgcolor{#1}} |
| 3824 |
% \end{macrocode} |
| 3825 |
% \end{macro} |
| 3826 |
% \end{macro} |
| 3827 |
% |
| 3828 |
% \begin{macro}{\bf@per@word} |
| 3829 |
% \begin{macro}{\KV@bytefield@per@word} |
| 3830 |
% Specify a macro to invoke for each word of a word box. The macro must |
| 3831 |
% take two arguments: the word number (0-indexed) and the total number |
| 3832 |
% of words. |
| 3833 |
% \begin{macrocode} |
| 3834 |
\define@key{bytefield}{perword}{\def\bf@per@word{#1}} |
| 3835 |
% \end{macrocode} |
| 3836 |
% \end{macro} |
| 3837 |
% \end{macro} |
| 3838 |
% |
| 3839 |
% \begin{macro}{\bytefieldsetup} |
| 3840 |
% \usermacro |
| 3841 |
% \changes{v2.0}{2011/01/18}{Introduced this macro to provide a more |
| 3842 |
% convenient way of configuring \pkgname{bytefield}'s parameters} |
| 3843 |
% \begin{macro}{\bf@bytefieldsetup} |
| 3844 |
% Reconfigure values for various \pkgname{bytefield} parameters. |
| 3845 |
% Internally to the package we use the |\bf@bytefieldsetup| macro |
| 3846 |
% instead of |\bytefieldsetup|. This enables us to redefine |
| 3847 |
% |\bytefieldsetup| when entering version~1 compatibility mode without |
| 3848 |
% impacting the rest of \pkgname{bytefield}. |
| 3849 |
% \begin{macrocode} |
| 3850 |
\newcommand{\bf@bytefieldsetup}{\setkeys{bytefield}} |
| 3851 |
\let\bytefieldsetup=\bf@bytefieldsetup |
| 3852 |
% \end{macrocode} |
| 3853 |
% \end{macro} |
| 3854 |
% \end{macro} |
| 3855 |
% |
| 3856 |
% We define only a single option that can be used only as a package |
| 3857 |
% option, not as an argument to |\bytefieldsetup|: |compat1| instructs |
| 3858 |
% \pkgname{bytefield} to enter version~1 compatibility mode---at the |
| 3859 |
% cost of a number of additional length registers and the inability to |
| 3860 |
% specify parameters in the argument to the |bytefield| environment. |
| 3861 |
% \begin{macrocode} |
| 3862 |
\DeclareOption{compat1}{\bf@enter@compatibility@mode@i} |
| 3863 |
% \end{macrocode} |
| 3864 |
% |
| 3865 |
% \begin{macro}{\bf@package@options} |
| 3866 |
% \begin{macro}{\next} |
| 3867 |
% We want to use |\bf@bytefieldsetup| to process \pkgname{bytefield} |
| 3868 |
% package options. Unfortunately, |\DeclareOption| doesn't handle |
| 3869 |
% \meta{key}=\meta{value} arguments. Hence, we use |\DeclareOption*| to |
| 3870 |
% catch \emph{all} options, each of which it appends to |
| 3871 |
% |\bf@package@options|. |\bf@package@options| is passed to |
| 3872 |
% |\bf@bytefieldsetup| only at the beginning of the document so that the |
| 3873 |
% options it specifies (a)~can refer to ex-heights and (b)~override the |
| 3874 |
% default values, which are also set at the beginning of the document. |
| 3875 |
% \begin{macrocode} |
| 3876 |
\def\bf@package@options{} |
| 3877 |
\DeclareOption*{% |
| 3878 |
\edef\next{% |
| 3879 |
\noexpand\g@addto@macro\noexpand\bf@package@options{,\CurrentOption}% |
| 3880 |
}% |
| 3881 |
\next |
| 3882 |
} |
| 3883 |
\ProcessOptions\relax |
| 3884 |
\expandafter\bf@bytefieldsetup\expandafter{\bf@package@options} |
| 3885 |
% \end{macrocode} |
| 3886 |
% \end{macro} |
| 3887 |
% \end{macro} |
| 3888 |
% |
| 3889 |
% \Finale |
| 3890 |
\endinput |