Debian Codegraph @

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

raw
1 % \iffalse meta-comment
2 %
3 % Copyright (C) 2011-2023 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 1.3c
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 2008/05/04 or later.
15 %
16 % \fi
17 %
18 % \iffalse
19 %<*driver>
20 \ProvidesFile{hyperxmp.dtx}
21 %</driver>
22 %<package>\NeedsTeXFormat{LaTeX2e}[1999/12/01]
23 %<package>\ProvidesPackage{hyperxmp}
24 %<*package>
25 [2024/03/17 v5.13 Store hyperref metadata in XMP format]
26 %</package>
27 %
28 %<*driver>
29 \bgroup
30 \catcode`\&=11
31 \gdef\levelchar{&}
32 \egroup
33 \documentclass{ltxdoc}
34 \usepackage[T1]{fontenc}
35 \usepackage[english]{babel}
36 \usepackage{graphicx}
37 \usepackage[cmyk]{xcolor}
38 \usepackage{tocbibind}
39 \usepackage{microtype}
40 \usepackage{needspace}
41 \usepackage{varioref}
42 \usepackage{fancyvrb}
43 \usepackage{multicol}
44 \usepackage{makeidx}
45 \usepackage{mmap}
46 \usepackage{enumitem}
47 \usepackage{listings}
48 \usepackage[labelformat=simple]{subcaption}
49 \renewcommand\thesubfigure{(\alph{subfigure})}
50 \usepackage{array}
51 \usepackage{booktabs}
52 \usepackage{marginfix}
53 \usepackage{ifmtarg}
54 \usepackage{iftex}
55 \input{hyperxmp-stds}
56 \expandafter\ifx\csname wantpdfstandards\endcsname\relax
57 \usepackage{hyperref}
58 \else
59 \usepackage[pdfa]{hyperref}
60 \fi
61 \usepackage{hyperxmp}
62 \EnableCrossrefs
63 \CodelineIndex
64 \RecordChanges
65
66 % Specify this document's metadata.
67 \GetFileInfo{hyperxmp.dtx}
68 \title{The \pkgname{hyperxmp} package\thanks{This document
69 corresponds to \pkgname{hyperxmp}~\fileversion, dated \filedate.}}
70 \author{Scott Pakin \\ \texttt{[email protected]}}
71 \makeatletter
72 \edef\versionnumber{\expandafter\@gobble\fileversion} % Drop the "v" in the file version.
73 \makeatother
74 \hypersetup{%
75 pdfauthor={Scott Pakin},
76 pdftitle={The hyperxmp package},
77 pdfsubject={LaTeX2e support for XMP metadata},
78 pdfkeywords={LaTeX, embedded metadata, XMP, PDF, copyright, license, comments},
79 pdfcopyright={Copyright (C) 2011-2023, Scott Pakin},
80 pdflicenseurl={http://www.latex-project.org/lppl/},
81 pdfcaptionwriter={Scott Pakin},
82 pdfcontactemail={[email protected]},
83 pdfcontacturl={http://www.pakin.org/\xmptilde scott/},
84 pdfversionid={\versionnumber},
85 pdflang={en-US},
86 pdftrapped={False},
87 pdfstartpage={},
88 pdfsource={hyperxmp.dtx},
89 pdfurl={http://mirror.ctan.org/macros/latex/contrib/hyperxmp/hyperxmp.pdf},
90 baseurl={http://mirror.ctan.org/macros/latex/contrib/hyperxmp/}
91 }
92
93 % Specify some additional hyperref configuration.
94 \hypersetup{%
95 unicode,
96 bookmarksopen,
97 bookmarksopenlevel=2,
98 bookmarksnumbered
99 }
100
101 % Redefine \changes to properly sort v5.10 *after* v5.9.
102 \makeatletter
103 \let\orig@changes=\changes
104 \def\discard@suffix#1\relax{}
105 \def\pad@minor@ver v#1.#2\relax{%
106 \afterassignment\discard@suffix
107 \@tempcnta=#2\relax
108 \edef\new@change@ver{%
109 v#1.%
110 \ifnum\@tempcnta<10
111 0#2
112 \else
113 #2
114 \fi
115 }%
116 }
117 \renewcommand{\changes}[2]{%
118 \pad@minor@ver#1\relax
119 \orig@changes{\new@change@ver=#1}{#2}%
120 }
121 \makeatother
122
123 \begin{document}
124 \DocInput{hyperxmp.dtx}
125 \Needspace{10\baselineskip}
126 \renewcommand*{\glossaryname}{Change History}
127 \PrintChanges
128 \Needspace{12\baselineskip}
129 \PrintIndex
130 \end{document}
131 %</driver>
132 % \fi
133 %
134 % \CheckSum{2859}
135 %
136 % \CharacterTable
137 % {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
138 % 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
139 % Digits \0\1\2\3\4\5\6\7\8\9
140 % Exclamation \! Double quote \" Hash (number) \#
141 % Dollar \$ Percent \% Ampersand \&
142 % Acute accent \' Left paren \( Right paren \)
143 % Asterisk \* Plus \+ Comma \,
144 % Minus \- Point \. Solidus \/
145 % Colon \: Semicolon \; Less than \<
146 % Equals \= Greater than \> Question mark \?
147 % Commercial at \@ Left bracket \[ Backslash \\
148 % Right bracket \] Circumflex \^ Underscore \_
149 % Grave accent \` Left brace \{ Vertical bar \|
150 % Right brace \} Tilde \~}
151 %
152 %
153 % \changes{v1.0}{2006/05/14}{Initial version}
154 % \changes{v2.0}{2012/08/02}{Heiko Oberdiek's major rewrite of the code
155 % to better support native-Unicode \string\TeX\ implementations
156 % (\string\XeTeX\ and \string\LuaTeX)}
157 % \changes{v2.0}{2012/08/26}{Added support for the \protect\acro{XMP} Basic
158 % schema and miscellaneous other bits of metadata}
159 % \changes{v1.2}{2010/06/04}{Added support for the \protect\XeTeX\ backend
160 % (\texttt{xdvipdfmx})}
161 % \changes{v1.2}{2010/06/07}{Added support for the Photoshop schema}
162 % \changes{v2.2}{2010/12/06}{Added support for the \protect\acro{IPTC} Photo Metadata schema}
163 % \changes{v2.4}{2013/12/21}{Added support for the \protect\acro{PDF/A}
164 % Identification schema, as requested by Florian Breitwieser}
165 % \changes{v2.5}{2014/06/19}{Enabled ``\texttt{\string\string\string\_}''
166 % to work within email addresses, as requested by Leonid Sinev}
167 % \changes{v2.6}{2014/09/24}{Added support for a new \protect\optname{pdfdate}
168 % key to explicitly specify the document date (and optionally time)}
169 % \changes{v2.9}{2016/04/25}{Force inclusion of \protect\xmpprop{dc:creator},
170 % \protect\xmpprop{dc:title}, and \protect\xmpprop{dc:description}---even if
171 % empty---when \protect\pkgname{hyperref} is loaded with the
172 % \protect\optname{pdfa} option (suggested by Leonid Sinev)}
173 % \changes{v2.9}{2016/04/26}{Introduced the \protect\optname{pdftype}
174 % package option, which enables an author to specify the type of
175 % document being produced}
176 % \changes{v3.0}{2016/07/02}{Made the code compatible with \string\LuaTeX~0.85.
177 % Thanks to Robert Schlicht, Leonid Sinev, and David Carlisle for bug
178 % reports and to Leonid Sinev for helping test the new
179 % \protect\pkgname{hyperxmp} code}
180 % \changes{v3.4}{2017/11/04}{Use \protect\pkgname{ifmtarg} to test for
181 % empty arguments, including non-empty but all spaces}
182 % \changes{v4.0}{2019/03/14}{Include all metadata within a single
183 % \protect\xmpterm{rdf:Description} block}
184 % \changes{v4.1}{2019/04/05}{Updated the documentation to refer to
185 % \protect\cs{pdfnumpages} by its correct name. Thanks to Volker RW Schaa
186 % for catching the discrepancy}
187 % \changes{v5.0}{2020/02/15}{Added support for \protect\acro{PDF/UA}
188 % standards, as requested by Robin Schwab}
189 % \changes{v5.0}{2020/02/26}{Added support for \protect\acro{PDF/X}
190 % standards, as requested by Robin Schwab}
191 % \changes{v5.0}{2020/03/17}{Don't set any document dates (creation,
192 % modification, or metadata) from \string\optname{pdfdate}}
193 % \changes{v5.2}{2020/04/29}{Introduced the \protect\optname{pdfidentifier}
194 % package option, which enables an author to specify a unique identifier
195 % for the document}
196 % \changes{v5.5}{2020/07/23}{Introduced the \protect\optname{pdfpubstatus}
197 % package option, which enables an author to specify the document's
198 % publication status. Thanks to Robin Schwab for pointing me to the
199 % Journal Article Versions recommendation~\protect\cite{NISO2008:jav}}
200 % \changes{v5.5}{2020/07/25}{Automatically assign
201 % \protect\optname{pdfnumpages} and \protect\optname{pdfbytes} under
202 % \protect\pdfLaTeX\ and \protect\LuaLaTeX}
203 % \changes{v5.5}{2020/09/23}{Defer \protect\cs{AtEndPreamble} execution
204 % until the end of the document. This enables
205 % \protect\pkgname{hyperxmp} itself to be loaded from
206 % \protect\cs{AtEndPreamble}, as is done by
207 % \protect\pkgname{doclicense}~v2.2.0. Thanks to Tommaso Pecorella
208 % for the bug report and help testing}
209 %
210 % ^^A \GetFileInfo{hyperxmp.dtx}
211 %
212 % \DoNotIndex{\,,\&,\ ,\!,\",\#,\(,\),\*,\<,\>,\\,\^,\~,\^,\_}
213 % \DoNotIndex{\@cons,\@empty,\@firstofone,\@firstoftwo,\@gobble,\@gobbletwo}
214 % \DoNotIndex{\@ifpackageloaded,\@ifundefined,\@nil,\@secondoftwo}
215 % \DoNotIndex{\@tempcnta,\@tempcntb,\MessageBreak,\advance,\afterassignment}
216 % \DoNotIndex{\aftergroup,\begin,\begingroup,\bgroup,\catcode,\csname,\def}
217 % \DoNotIndex{\divide,\do,\edef,\egroup,\else,\end,\endcsname,\endgroup}
218 % \DoNotIndex{\expandafter,\fi,\futurelet,\g@addto@macro,\gdef,\global}
219 % \DoNotIndex{\hbox,\if,\ifcase,\ifnum,\ifx,\immediate,\lccode,\let,\loop}
220 % \DoNotIndex{\lowercase,\multiply,\newcommand,\noexpand,\or,\relax,\repeat}
221 % \DoNotIndex{\space,\string,\the,\toks,\uccode,\uppercase,\usepackage}
222 % \DoNotIndex{\xdef}
223 %
224 % ^^A Define a few logical styles.
225 % \DeclareRobustCommand{\term}[1]{#1\SortIndex{#1}{#1}}
226 % \DeclareRobustCommand{\pkgname}[1]{^^A
227 % \mbox{\textsf{#1}}^^A
228 % \SortIndex{#1}{\textsf{#1} (package)}^^A
229 % \index{packages&#1=\textsf{#1}}^^A
230 % }
231 % \DeclareRobustCommand{\clsname}[1]{^^A
232 % \mbox{\textsf{#1}}^^A
233 % \SortIndex{#1}{\textsf{#1} (class)}^^A
234 % \index{classes&#1=\textsf{#1}}^^A
235 % }
236 % \makeatletter
237 % \DeclareRobustCommand{\xmpprop}[2][]{^^A XMP property
238 % \def\xmppropopt{#1}^^A
239 % \ifx\xmppropopt\@empty
240 % \mbox{\textsf{#2}}^^A
241 % \else
242 % \mbox{\textsf{#2}}.\mbox{\textsf{#1}}^^A
243 % \SortIndex{#1}{\textsf{#1}}^^A
244 % \fi
245 % \SortIndex{#2}{\textsf{#2}}^^A
246 % \index{properties, \acrostyle{XMP}&#2=\textsf{#2}}^^A
247 % }
248 % \DeclareRobustCommand{\xmpterm}[1]{^^A All other XMP terms
249 % \mbox{\textsf{#1}}^^A
250 % \SortIndex{#1}{\textsf{#1}}^^A
251 % }
252 % \makeatother
253 % \index{XMP=\acrostyle {XMP}&properties|see{properties, \acrostyle{XMP}}}
254 % \DeclareRobustCommand{\pdfterm}[1]{^^A
255 % \mbox{\textsf{#1}}^^A
256 % \SortIndex{#1}{\textsf{#1}}^^A
257 % \index{PDF=\acrostyle{PDF}&#1=\textsf{#1}}^^A
258 % }
259 % \DeclareRobustCommand{\cmdname}[1]{\mbox{\texttt{#1}}\SortIndex{#1}{\texttt{#1}}}
260 % \DeclareRobustCommand{\optname}[1]{^^A
261 % \mbox{\textsf{#1}}^^A
262 % \SortIndex{#1}{\textsf{#1} (option)}^^A
263 % \index{options&#1=\textsf{#1}}^^A
264 % }
265 % ^^A \moptname is the same as \optname but additionally typesets the
266 % ^^A option name in the margin.
267 % \DeclareRobustCommand{\moptname}[1]{^^A
268 % \optname{#1}^^A
269 % \marginpar{\raggedleft\strut\mbox{\textsf{#1}}}^^A
270 % }
271 % \DeclareRobustCommand{\acrostyle}[1]{\textsc{\MakeLowercase{#1}}}
272 % \DeclareRobustCommand{\acro}[1]{^^A
273 % \mbox{\acrostyle{#1}}^^A
274 % \SortIndex{#1}{\acrostyle{#1}}^^A
275 % }
276 % \newcounter{note}
277 % \newcommand{\usagenote}[1]{^^A
278 % \refstepcounter{note}^^A
279 % \Needspace{\baselineskip}^^A
280 % \belowpdfbookmark{Note \thenote: #1}{note.\thenote}^^A
281 % \paragraph{Note~\thenote: #1}%
282 % }
283 %
284 % ^^A Define some other shortcut macros.
285 % \DeclareRobustCommand{\tex}{^^A
286 % \texorpdfstring{\TeX\SortIndex{TeX}{\TeX}}{TeX}^^A
287 % }
288 % \DeclareRobustCommand{\XeTeXlogo}{^^A
289 % X\lower0.5ex\hbox{\kern-.15em\reflectbox{E}}\kern-0.1667em\TeX
290 % }
291 % \DeclareRobustCommand{\XeTeX}{^^A
292 % \texorpdfstring{\XeTeXlogo\SortIndex{XeTeX}{\XeTeXlogo}}{XeTeX}^^A
293 % }
294 % \DeclareRobustCommand{\XeLaTeXlogo}{^^A
295 % X\lower0.5ex\hbox{\kern-.15em\reflectbox{E}}\kern-0.1667em\LaTeX
296 % }
297 % \DeclareRobustCommand{\XeLaTeX}{^^A
298 % \texorpdfstring{\XeLaTeXlogo\SortIndex{XeLaTeX}{\XeLaTeXlogo}}{XeLaTeX}^^A
299 % }
300 % \DeclareRobustCommand{\LuaTeXlogo}{Lua\TeX}
301 % \DeclareRobustCommand{\LuaTeX}{^^A
302 % \texorpdfstring{\LuaTeXlogo\SortIndex{LuaTeX}{\LuaTeXlogo}}{LuaTeX}^^A
303 % }
304 % \DeclareRobustCommand{\LuaLaTeXlogo}{Lua\LaTeX}
305 % \DeclareRobustCommand{\LuaLaTeX}{^^A
306 % \texorpdfstring{\LuaLaTeXlogo\SortIndex{LuaLaTeX}{\LuaLaTeXlogo}}{LuaLaTeX}^^A
307 % }
308 % \DeclareRobustCommand{\pdfTeXlogo}{pdf\TeX}
309 % \DeclareRobustCommand{\pdfTeX}{^^A
310 % \texorpdfstring{\pdfTeXlogo\SortIndex{pdfTeX}{\pdfTeXlogo}}{pdfTeX}^^A
311 % }
312 % \DeclareRobustCommand{\pdfLaTeXlogo}{pdf\LaTeX}
313 % \DeclareRobustCommand{\pdfLaTeX}{^^A
314 % \texorpdfstring{\pdfLaTeXlogo\SortIndex{pdfLaTeX}{\pdfLaTeXlogo}}{pdfLaTeX}^^A
315 % }
316 % \DeclareRobustCommand{\Dvips}{^^A
317 % \texorpdfstring{Dvips\SortIndex{dvips}{\texttt{dvips}}}{Dvips}^^A
318 % }
319 % \DeclareRobustCommand{\progname}[1]{^^A
320 % \mbox{\texttt{#1}}\SortIndex{#1}{\texttt{#1}}^^A
321 % }
322 % \DeclareRobustCommand{\Lua}{^^A
323 % Lua\index{Lua}^^A
324 % }
325 % \DeclareRobustCommand{\Koma}{^^A
326 % Koma^^A
327 % \SortIndex{Koma}{Koma (class)}^^A
328 % \index{classes&Koma}^^A
329 % }
330 % \DeclareRobustCommand{\ACMclass}{^^A
331 % \acro{ACM}^^A
332 % \index{classes&ACM=\acrostyle{ACM}}^^A
333 % }
334 %
335 % ^^A Define an environment just like macro but for Lua functions.
336 % \makeatletter
337 % \newenvironment{luafunc}{^^A
338 % \begingroup
339 % \catcode`\_=11
340 % \def\PrintMacroName##1{\strut\MacroFont\string##1\ }^^A
341 % \def\SpecialIndex@##1##2{^^A
342 % \@bsphack
343 % \special@index{\string##1\actualchar
344 % \string\verb\quotechar*\verbatimchar##1\verbatimchar##2}^^A
345 % \@esphack
346 % }^^A
347 % \begin{macro}^^A
348 % }{^^A
349 % \end{macro}^^A
350 % \endgroup
351 % }
352 % \makeatother
353 %
354 % ^^A Typeset a PDF standard name, indexing only the standard type but
355 % ^^A not the subtype or year.
356 % \makeatletter
357 % \newcommand*{\PDFstd}[4]{^^A
358 % \mbox{^^A
359 % \acro{PDF/#1}-\oldstylenums{#2}^^A
360 % #3^^A
361 % \@ifnotmtarg{#4}{:\oldstylenums{#4}}^^A
362 % }^^A
363 % }
364 % \makeatother
365 %
366 % ^^A Index schemata.
367 % \DeleteShortVerb{\|}
368 % \makeatletter
369 % \newenvironment{schemadesc}[2][]{^^A
370 % \@ifmtarg{#1}{^^A
371 % \index{#2 schema|(}^^A
372 % \index{schemata&#2|(}^^A
373 % \def\closeschemedesc{^^A
374 % \index{schemata&#2|)}^^A
375 % \index{#2 schema|)}^^A
376 % }^^A
377 % }{^^A
378 % \index{#1 schema=#2 schema|(}^^A
379 % \index{schemata&#1=#2|(}^^A
380 % \def\closeschemedesc{^^A
381 % \index{schemata&#1=#2|)}^^A
382 % \index{#1 schema=#2 schema|)}^^A
383 % }^^A
384 % }^^A
385 % }{^^A
386 % \closeschemedesc
387 % }
388 % \newcommand{\schema}[2][]{^^A
389 % \@ifmtarg{#1}{^^A
390 % \kern0pt^^A
391 % \index{#2 schema}^^A
392 % \index{schemata&#2}^^A
393 % #2^^A
394 % }{^^A
395 % \kern0pt^^A
396 % \index{#1 schema=#2 schema}^^A
397 % \index{schemata&#1=#2}^^A
398 % #2^^A
399 % }^^A
400 % }
401 % \makeatother
402 % \MakeShortVerb{\|}
403 %
404 % ^^A Pack figures a bit tighter onto the page.
405 % \renewcommand{\floatpagefraction}{0.8}
406 %
407 % ^^A Help \pageref refer to arbitrary content.
408 % \newcounter{pagelabel}
409 %
410 % \maketitle
411 % \sloppy
412 %
413 % \begin{abstract}
414 % \pkgname{hyperxmp} makes it easy for an author to include \acro{XMP}
415 % metadata in a \acro{PDF} document produced by \LaTeX\@. \pkgname{hyperxmp}
416 % integrates seamlessly with \pkgname{hyperref} and requires virtually
417 % no modifications to a document that already specifies document
418 % metadata through \pkgname{hyperref}'s mechanisms.
419 % \end{abstract}
420 %
421 % \section{Introduction}
422 %
423 % Adobe Systems, Inc.\ has been promoting
424 % \acro{XMP}~\cite{Adobe2012:XMP}---eXtensible Metadata Platform---as a
425 % standard way to include metadata within a document. The idea behind
426 % \acro{XMP} is that it is an \acro{XML}-based description of various
427 % document attributes and is embedded as uncompressed, unencoded text
428 % within the document it describes. By storing the metadata this way it
429 % is independent of the document's file format. That is, regardless of
430 % whether a document is in \acro{PDF}, \acrostyle{JPEG},
431 % \acrostyle{HTML}, or any other format, it is trivial for a program (or
432 % human) to locate, extract, and---using any standard \acro{XML}
433 % parser---process the embedded \acro{XMP} metadata.
434 %
435 % As of this writing there are few tools that actually do process
436 % \acro{XMP}\@. However, it is easy to imagine future support existing
437 % in file browsers for displaying not only a document's filename but
438 % also its title, list of authors, description, and other metadata.
439 %
440 % \paragraph{This is too abstract! Give me an example.}
441 % Consider a \LaTeX\ document with three authors---Jack Napier, Edward
442 % Nigma, and Harvey Dent---named in the \LaTeX\ source in the usual way:
443 % ``\verb|\author{|\texttt{Jack Napier \cs{and} Edward Nigma \cs{and}
444 % Harvey Dent}\verb|}|''. With \pkgname{hyperxmp}, the generated
445 % \acro{PDF} file will contain, among other information, the following
446 % stanza of \acro{XMP} code embedded within it:
447 %
448 % \begin{verbatim}
449 % <dc:creator>
450 % <rdf:Seq>
451 % <rdf:li>Jack Napier</rdf:li>
452 % <rdf:li>Edward Nigma</rdf:li>
453 % <rdf:li>Harvey Dent</rdf:li>
454 % </rdf:Seq>
455 % </dc:creator>
456 % \end{verbatim}
457 %
458 % In the preceding code, the |dc| namespace refers to the
459 % \href{http://purl.org/DC/}{\schema{Dublin Core} schema}, a collection
460 % of metadata properties. The \xmpprop{dc:creator} property surrounds
461 % the list of authors. The \textsf{rdf} namespace is the
462 % \href{http://www.w3.org/RDF/}{Resource Description Framework}, which
463 % defines \xmpterm{rdf:Seq} as an ordered list of values. Each author
464 % is represented by an individual list item (\xmpterm{rdf:li}), making
465 % it easy for an \acro{XML} parser to separate the authors' names.
466 %
467 % Remember that \acro{XMP} code is stored as \emph{metadata}. It does
468 % not appear when viewing or printing the \acro{PDF} file. Rather, it
469 % is intended to make it easy for computer applications to identify and
470 % categorize the document.
471 %
472 % \subsection{Supported metadata}
473 %
474 % \pkgname{hyperxmp} knows how to embed all of the following types of
475 % metadata within a document:
476 %
477 % \label{page:begin-metadata-list}
478 % \begin{itemize} \raggedright
479 % \item address of primary author
480 % (\xmpprop[CiAdrExtadr]{Iptc4xmpCore:CreatorContactInfo},
481 % \xmpprop[CiAdrCity]{Iptc4xmpCore:CreatorContactInfo},
482 % \xmpprop[CiAdrRegion]{Iptc4xmpCore:CreatorContactInfo},
483 % \xmpprop[CiAdrPcode]{Iptc4xmpCore:CreatorContactInfo}, and
484 % \xmpprop[CiAdrCtry]{Iptc4xmpCore:CreatorContactInfo})
485 % \item author(s) (\xmpprop{dc:creator})
486 % \item base \acro{URL} for relative references (\xmpprop{xmp:BaseURL})
487 % \item book edition (\xmpprop{prism:bookEdition})
488 % \item copyright (\xmpprop{dc:rights} and \xmpprop{xmpRights:Marked})
489 % \item date (\xmpprop{dc:date}, \xmpprop{xmp:CreateDate},
490 % \xmpprop{xmp:ModifyDate}, and \xmpprop{xmp:MetadataDate})
491 % \item \acro{DOI} (\xmpprop{prism:doi})
492 % \item email address(es) of primary author (\xmpprop[CiEmailWork]{Iptc4xmpCore:CreatorContactInfo})
493 % \item file format (\xmpprop{dc:format})
494 % \item file name of main \LaTeX\ source file (\xmpprop{dc:source})
495 % \item file size in bytes (\xmpprop{prism:byteCount})
496 % \item \acro{ISBN} (\xmpprop{prism:isbn})
497 % \item \acro{ISSN}---both print (\xmpprop{prism:issn}) and electronic (\xmpprop{prism:eIssn})
498 % \item issue number of parent publication (\xmpprop{prism:number})
499 % \item journal article version (\xmpprop{jav:journal\_article\_version})
500 % \item keywords (\xmpprop{pdf:Keywords} and \xmpprop{dc:subject})
501 % \item language used (\xmpprop{dc:language})
502 % \item license \acro{URL} (\xmpprop{xmpRights:WebStatement})
503 % \item metadata writer (\xmpprop{photoshop:CaptionWriter})
504 % \item page count (\xmpprop{prism:pageCount})
505 % \item page range(s) (\xmpprop{prism:pageRange})
506 % \item \acro{PDF} version (\xmpprop{pdf:PDFVersion})
507 % \item \acro{PDF}-generating tool (\xmpprop{pdf:Producer} and \xmpprop{xmp:CreatorTool})
508 % \item \acro{PDF/A} version and conformance level (\xmpprop{pdfaid:part} and \xmpprop{pdfaid:conformance})
509 % \item \acro{PDF/UA} version (\xmpprop{pdfuaid:part})
510 % \item \acro{PDF/X} standard compliance (\xmpprop{pdfxid:GTS\_PDFXVersion})
511 % \item position/title of primary author (\xmpprop{photoshop:AuthorsPosition})
512 % \item publication name of parent publication (\xmpprop{prism:publicationName})
513 % \item publisher of the document (\xmpprop{dc:publisher})
514 % \item rendition variation of the document (\xmpprop{xmpMM:RenditionClass})
515 % \item summary (\xmpprop{dc:description})
516 % \item subtitle (\xmpprop{prism:subtitle})
517 % \item telephone number(s) of primary author (\xmpprop[CiTelWork]{Iptc4xmpCore:CreatorContactInfo})
518 % \item title (\xmpprop{dc:title})
519 % \item trapping of colors (\xmpprop{pdf:trapped})
520 % \item type of document (\xmpprop{dc:type})
521 % \item type of parent publication (\xmpprop{prism:aggregationType})
522 % \item unique identifier for the document (\xmpprop{dc:identifier})
523 % \item \acro{URL} of the document (\xmpprop{prism:url})
524 % \item \acro{URL}(s) of the primary author (\xmpprop[CiUrlWork]{Iptc4xmpCore:CreatorContactInfo})
525 % \item \acro{UUID} for the document (\xmpprop{xmpMM:DocumentID})
526 % \item \acro{UUID} for the document instance (\xmpprop{xmpMM:InstanceID})
527 % \item version identifier for the document (\xmpprop{xmpMM:VersionID})
528 % \item volume number of parent publication (\xmpprop{prism:volume})
529 % \end{itemize}
530 % \label{page:end-metadata-list}
531 %
532 % \noindent
533 % More types of metadata may be added in a future release.
534 %
535 % \subsection{Comparisons with similar packages}
536 %
537 % \paragraph{\textsf{xmpincl}}
538 % In short, \pkgname{xmpincl} is more flexible but \pkgname{hyperxmp} is
539 % easier to use. With \pkgname{xmpincl}, the author manually constructs
540 % a file of arbitrary \acro{XMP} data and the package merely embeds it
541 % within the generated \acro{PDF} file. With \pkgname{hyperxmp}, the
542 % author specifies values for various predefined metadata types and the
543 % package formats those values as \acro{XMP} and embeds the result
544 % within the generated \acro{PDF} file.
545 %
546 % \pkgname{xmpincl} can embed \acro{XMP} only when running under \pdfLaTeX\ and
547 % only when in \acro{PDF}-generating mode. \pkgname{hyperxmp} additionally
548 % works with a few other \acro{PDF}-producing \LaTeX\ backends.
549 %
550 % \pkgname{hyperxmp} and \pkgname{xmpincl} can complement each other.
551 % An author may want to use \pkgname{hyperxmp} to produce a basic set of
552 % \acro{XMP} code, then extract the \acro{XMP} code from the \acro{PDF} file with a text
553 % editor, augment the \acro{XMP} code with any metadata not supported by
554 % \pkgname{hyperxmp}, and use \pkgname{xmpincl} to include the modified
555 % \acro{XMP} code in the \acro{PDF} file.
556 %
557 % \paragraph{\textsf{pdfx}}
558 % The main difference between \pkgname{hyperxmp} and \pkgname{pdfx} is
559 % that \pkgname{hyperxmp} tries to integrate as seamlessly as possible
560 % into an existing document. It leverages \pkgname{hyperref}'s
561 % \cs{hypersetup} command and many of \cs{hypersetup}'s options and
562 % defines its own options in a compatible manner. In contrast,
563 % \pkgname{pdfx} requires the user to create a separate
564 % |\jobname.xmpdata| file containing \pkgname{pdfx}-defined commands for
565 % each metadata element.
566 %
567 % \iffalse
568 %<*listings>
569 % \fi
570 \newsavebox{\pdfxbox}
571 \begin{lrbox}{\pdfxbox}
572 \begin{lstlisting}[
573 language=tex,
574 columns=fullflexible,
575 linewidth=14pc,
576 frame=single,
577 basicstyle=\footnotesize,
578 showlines,
579 morekeywords={\Title,\Author,\Language,\Keywords,\Publisher,\sep}
580 ]
581
582 \Title{Baking through the ages}
583 \Author{A. Baker\sep C. Kneader}
584 \Language{en-GB}
585 \Keywords{cookies\sep muffins\sep cakes}
586 \Publisher{Baking International}
587
588 \end{lstlisting}
589 \end{lrbox}
590
591 \newsavebox{\hyxmpbox}
592 \begin{lrbox}{\hyxmpbox}
593 \begin{lstlisting}[
594 language=tex,
595 columns=fullflexible,
596 linewidth=13pc,
597 frame=single,
598 basicstyle=\footnotesize,
599 morekeywords={\hypersetup}
600 ]
601 \hypersetup{%
602 pdftitle={Baking through the ages},
603 pdfauthor={A. Baker, C. Kneader},
604 pdflang={en-GB},
605 pdfkeywords={cookies, muffins, cakes},
606 pdfpublisher={Baking International}
607 }
608 \end{lstlisting}
609 \end{lrbox}
610 % \iffalse
611 %</listings>
612 % \fi
613 %
614 % Figure~\ref{fig:pdfx-hyxmp} adapts an example appearing in the
615 % \pkgname{pdfx} manual to \pkgname{hyperxmp}.
616 % \begin{figure}
617 % \centering
618 % \subcaptionbox{\pkgname{pdfx} (separate \texttt{.xmpdata} file)
619 % \label{fig:ph-pdfx}}{%
620 % \usebox{\pdfxbox}%
621 % }
622 % \hfil
623 % \subcaptionbox{\pkgname{hyperxmp} (main document)
624 % \label{fig:ph-hyxmp}}{%
625 % \usebox{\hyxmpbox}}
626 % \caption{Comparison of \pkgname{pdfx} and \pkgname{hyperxmp}}
627 % \label{fig:pdfx-hyxmp}
628 % \end{figure}
629 % The two are comparable line-by-line in terms of how one specifies the
630 % title, author, document language, keywords, and publisher. However,
631 % \pkgname{hyperxmp} implicitly writes a wealth of additional metadata
632 % into the \acro{XMP} packet such as the document date, creation date,
633 % creator tool, file format, \acro{PDF} version, and unique document and
634 % instance IDs. In fact, if a document omits all of the code shown in
635 % Figure~\ref{fig:ph-hyxmp}, it will still store the \cs{title} and
636 % \cs{author} data in the \acro{XMP} packet.
637 %
638 % One can therefore summarize the difference between \pkgname{hyperxmp}
639 % and \pkgname{pdfx} as follows: \pkgname{pdfx} requires the author to
640 % be fully explicit about the document's metadata while
641 % \pkgname{hyperxmp} allows some metadata to be specified implicitly,
642 % automatically inferring it when possible. In general,
643 % \pkgname{hyperxmp} tries to simplify the author's task as much as
644 % possible.
645 %
646 %
647 % \section{Usage}
648 %
649 % \pkgname{hyperxmp} works by postprocessing some of the package options
650 % honored by \pkgname{hyperref}. To use \pkgname{hyperxmp}, merely put
651 % a |\usepackage{hyperxmp}| in your document's preamble. That line can
652 % appear anywhere \emph{after} the |\usepackage{hyperref}| but
653 % \emph{before} \pkgname{hyperref}'s \acro{PDF} options are specified
654 % with |\hypersetup|. \pkgname{hyperxmp} will construct its \acro{XMP}
655 % data using the following \pkgname{hyperref} options:
656 %
657 % \begin{multicols}{3}
658 % \raggedcolumns
659 % \begin{itemize}
660 % \item \optname{baseurl}
661 % \item \optname{pdfauthor}
662 % \item \optname{pdfcreationdate}
663 % \item \optname{pdfkeywords}
664 % \item \optname{pdflang}
665 % \item \optname{pdfmoddate}
666 % \item \optname{pdfproducer}
667 % \item \optname{pdfsubject}
668 % \item \optname{pdftitle}
669 % \item \optname{pdftrapped}
670 % \end{itemize}
671 % \end{multicols}
672 %
673 % \noindent
674 % \mbox{}\label{page:begin-new-options}
675 % \pkgname{hyperxmp} instructs \pkgname{hyperref} also to accept the
676 % following options, which have meaning only to \pkgname{hyperxmp}:
677 %
678 % \begin{multicols}{3}
679 % \raggedcolumns
680 % \begin{itemize}
681 % \item \optname{pdfaconformance}
682 % \item \optname{pdfapart}
683 % \item \optname{pdfauthortitle}
684 % \item \optname{pdfbookedition}
685 % \item \optname{pdfbytes}
686 % \item \optname{pdfcaptionwriter}
687 % \item \optname{pdfcontactaddress}
688 % \item \optname{pdfcontactcity}
689 % \item \optname{pdfcontactcountry}
690 % \item \optname{pdfcontactemail}
691 % \item \optname{pdfcontactphone}
692 % \item \optname{pdfcontactpostcode}
693 % \item \optname{pdfcontactregion}
694 % \item \optname{pdfcontacturl}
695 % \item \optname{pdfcopyright}
696 % \item \optname{pdfdate}
697 % \item \optname{pdfdocumentid}
698 % \item \optname{pdfdoi}
699 % \item \optname{pdfeissn}
700 % \item \optname{pdfidentifier}
701 % \item \optname{pdfinstanceid}
702 % \item \optname{pdfisbn}
703 % \item \optname{pdfissn}
704 % \item \optname{pdfissuenum}
705 % \item \optname{pdflicenseurl}
706 % \item \optname{pdfmetadate}
707 % \item \optname{pdfmetalang}
708 % \item \optname{pdfnumpages}
709 % \item \optname{pdfpagerange}
710 % \item \optname{pdfpublication}
711 % \item \optname{pdfpublisher}
712 % \item \optname{pdfpubstatus}
713 % \item \optname{pdfpubtype}
714 % \item \optname{pdfrendition}
715 % \item \optname{pdfsource}
716 % \item \optname{pdfsubtitle}
717 % \item \optname{pdftype}
718 % \item \optname{pdfuapart}
719 % \item \optname{pdfurl}
720 % \item \optname{pdfversionid}
721 % \item \optname{pdfvolumenum}
722 % \item \optname{pdfxstandard}
723 % \end{itemize}
724 % \end{multicols}
725 % \label{page:end-new-options}
726 %
727 %
728 % \subsection{Option descriptions}
729 %
730 % The document title is specified as normal for \pkgname{hyperref} with
731 % \moptname{pdftitle}, but see Note~\ref{note:multilingual} on
732 % page~\pageref{note:multilingual} for instructions on how to specify a
733 % title in multiple languages. If \optname{pdftitle} is not specified
734 % it will inherit its value from the document's \cs{title}.
735 % \pkgname{hyperxmp} introduces a complementary \moptname{pdfsubtitle}
736 % option:
737 %
738 % \begin{verbatim}
739 % pdftitle={Frankenstein},
740 % pdfsubtitle={The Modern Prometheus},
741 % \end{verbatim}
742 %
743 % Unfortunately, the subtitle can appear in only one language. It
744 % assumed to be the same language as the document language
745 % (\optname{pdflang}) but can be overridden by preceding the text with a
746 % bracketed \acro{ISO} \mbox{639-1} two-letter language code and an
747 % optional \acro{ISO} \mbox{3166-1} two-letter region code. See the
748 % example below for \optname{pdfpublication}.
749 %
750 % \pkgname{hyperref}'s \moptname{pdfauthor} option specifies the
751 % document's author(s). See Note~\ref{note:literal-commas} on
752 % page~\pageref{note:literal-commas} for a discussion of the correct
753 % syntax. If \optname{pdfauthor} is not specified it will inherit its
754 % value from the document's \cs{author}. \moptname{pdfauthortitle}
755 % indicates the primary author's position or title.
756 % \moptname{pdfcaptionwriter} specifies the name of the person who added
757 % the metadata to the document.
758 %
759 % The next eight items describe how to contact the person or institution
760 % responsible for the document (the ``contact'').
761 % \moptname{pdfcontactaddress} is the contact's street address and can
762 % include the institution name if the contact is an institution;
763 % \moptname{pdfcontactcity} is the contact's city;
764 % \moptname{pdfcontactcountry} is the contact's country;
765 % \moptname{pdfcontactemail} is the contact's email address (or
766 % multiple, comma-separated email addresses); \moptname{pdfcontactphone}
767 % is the contact's telephone number (or multiple, comma-separated
768 % telephone numbers); \moptname{pdfcontactpostcode} is the contact's
769 % postal code; \moptname{pdfcontactregion} is the contact's state or
770 % province; and \moptname{pdfcontacturl} is the contact's \acro{URL} (or
771 % multiple, comma-separated \acro{URL}s).
772 %
773 % \Needspace{2\baselineskip}
774 % \moptname{pdfcopyright} defines the copyright text, and
775 % \moptname{pdflicenseurl} identifies a \acro{URL} that points to the
776 % document's license agreement.
777 %
778 % \moptname{pdfmetalang} indicates the natural language in which certain
779 % metadata---specifically, the document's title, subject, and copyright
780 % statement---are written. The language should be specified using an
781 % \acro{IETF} language tag~\cite{IANA2011:lang-tags}, for example,
782 % ``|en|'' for English, ``|en-US|'' for specifically United States
783 % English, ``|de|'' for German, and so forth. If \optname{pdfmetalang}
784 % is not specified, \pkgname{hyperxmp} assumes the metadata language is
785 % the same as the document language (\pkgname{hyperref}'s
786 % \moptname{pdflang} option). If neither \optname{pdfmetalang} nor
787 % \optname{pdflang} is specified, \pkgname{hyperxmp} uses only
788 % ``\xmpterm{x-default}'' as the metadata language.
789 %
790 % \acro{XMP} can include a universally unique identifier (\acro{UUID})
791 % for each document and for each instance of a given document. By
792 % default, \pkgname{hyperxmp} assigns a version~4 (i.e.,~pseudorandom)
793 % \acro{UUID}~\cite{Leach2005:uuid} for each of these. However, a
794 % document can alternatively specify a particular document identifier
795 % using \moptname{pdfdocumentid} and (not normally recommended) a
796 % particular instance identifier using \moptname{pdfinstanceid}. These
797 % should be of the form
798 % \texttt{uuid:}\textit{xxxxxxxx}\texttt{-}\textit{xxxx}\texttt{-}\textit{xxxx}\texttt{-}\textit{xxxx}\texttt{-}\textit{xxxxxxxxxxxx},
799 % where ``\textit{x}'' is a lowercase hexadecimal number. For example,
800 % \texttt{uuid:53ab7f19-a48c-5177-8bb2-403ad907f632} is a valid argument
801 % to \optname{pdfdocumentid} (or \optname{pdfinstanceid}). See Leach,
802 % Mealling, and Salz's \acro{UUID} specification document for details on
803 % how to produce the various forms of
804 % \acro{UUID}s~\cite{Leach2005:uuid}. A more freeform mechanism than
805 % \optname{pdfinstanceid} for versioning documents is available via
806 % \moptname{pdfversionid}. The version specified by
807 % \optname{pdfversionid} can be incremented as~|1|, |2|, |3|,~\dots;
808 % identified with a hierarchical numbering scheme (e.g.,~this document
809 % is versioned \texttt{\versionnumber} to match the package version); or
810 % labeled using any other approach. One possibility is to use a
811 % revision number or commit hash from the version-control software
812 % maintaining the document. For example, the |\gitVer| macro from the
813 % \pkgname{gitver} package is an expandable (see
814 % Note~\ref{note:expandable} on page~\pageref{note:expandable}) version
815 % of the current \href{https://git-scm.com/}{Git} hash that can suitably
816 % be passed to \optname{pdfversionid}. If not specified,
817 % \optname{pdfversionid} defaults to~|1|.
818 %
819 % Already-published documents can be identified in a number of ways.
820 % \moptname{pdfisbn} specifies the \acro{ISBN}. \moptname{pdfissn}
821 % refers to the \acro{ISSN} of the \emph{print} version of the document
822 % while \moptname{pdfeissn} refers to the \acro{ISSN} of the
823 % \emph{electronic} version of the document. \moptname{pdfdoi}
824 % specifies the \acro{DOI} and should include only the \acro{DOI} name
825 % without any \acro{URL} prefix. For example, specify
826 % |pdfdoi={10.1145/3149526.3149532}|, \emph{not}
827 % |pdfdoi={https://doi.org/10.1145/3149526.3149532}|.
828 % \moptname{pdfurl} points to the complete \acro{URL} for the document.
829 % In contrast, \moptname{baseurl} points one level up and is used to
830 % resolve relative \acro{URL}s.
831 %
832 % \moptname{pdfidentifier} provides an alternative mechanism to uniquely
833 % identify a document. Its advantage relative to \optname{pdfisbn},
834 % \optname{pdfissn}, \optname{pdfdoi}, etc.\ is its flexibility; any of
835 % a wide variety of identification types can be used.\footnote{See, for
836 % example,
837 % \url{https://www.iana.org/assignments/urn-namespaces/urn-namespaces.xhtml}
838 % for the \texttt{urn:} \acro{URI} scheme and
839 % \url{http://info-uri.info/registry/} for the \texttt{info:}
840 % \acro{URI} scheme.} \optname{pdfidentifier}'s disadvantage is that
841 % it allows only a single identifier per document. For example, a
842 % document could use |pdfidentifier=urn:iso:std:32000:ed-1:v1:en| to
843 % identify itself as version~1 of English-language \acro{ISO} standard
844 % 32000-1, but then this same document could not also use
845 % \optname{pdfidentifier} to identify itself by \acro{DOI}
846 % (|info:doi/|\dots), \acro{ISBN} (|urn:ISSN:|\dots), etc. (It can
847 % still use the options described in the previous paragraph, though.) If
848 % \optname{pdfidentifier} is not specified explicitly,
849 % \pkgname{hyperxmp} will use the first non-empty value out of the
850 % \acro{DOI}, electronic \acro{ISSN}, print \acro{ISSN}, and \acro{ISBN}
851 % or skip the identifier entirely if all of those are empty.
852 %
853 % Already-published documents can further be identified by the
854 % publication in which they appear. \moptname{pdfpublication} specifies
855 % the title of the journal, magazine, or other parent document. The
856 % title language is assumed to be the same as the document language
857 % (\optname{pdflang}) but can be overridden by preceding the text with a
858 % bracketed \acro{ISO} \mbox{639-1} two-letter language code and an
859 % optional \acro{ISO} \mbox{3166-1} two-letter region code. For
860 % example, |pdfpublication={[fr]Charlie Hedbo}| indicates a
861 % French-language title. Were the language or pronunciation differences
862 % significant, |fr-FR| would indicate specifically the French spoken in
863 % France, as opposed to that spoken in, say, Canada (|fr-CA|) or Belgium
864 % (|fr-BE|). The publisher itself can be named using
865 % \moptname{pdfpublisher}.
866 %
867 % \moptname{pdfpubtype} indicates the type of publication in which the
868 % document was published. This should be one of the \acro{PRISM}
869 % aggregation types~\cite{PRISM2012:cont-voc} such as |book|, |journal|,
870 % |magazine|, |manual|, |report|, or |whitepaper|.
871 %
872 % For publications in journals, magazines, and similar periodicals, a
873 % document can specify the volume number with \moptname{pdfvolumenum}
874 % and the issue number within the volume with \moptname{pdfissuenum}.
875 % \moptname{pdfpagerange} indicates the page numbers at which the
876 % document appears within the publication. The intention is that this
877 % be a comma-separated list of dash-separated ranges, as in
878 % |pdfpagerange={1,4-5}|. See Note~\ref{note:page-counting} on
879 % page~\pageref{note:page-counting} for advice on how to assign
880 % \optname{pdfpagerange} semi-automatically. A journal article's
881 % publication status can be indicated with \moptname{pdfpubstatus}.
882 % This option expects to take one of the values listed in
883 % Table~\ref{tbl:jav}. See the \acro{NISO}/\acro{ALPSP} Journal Article
884 % Versions recommendation~\cite{NISO2008:jav} for an explanation of each
885 % of those values and when to use them.
886 %
887 % \begin{table}
888 % \centering
889 % \caption{Valid arguments for \optname{pdfpubstatus}}
890 % \label{tbl:jav}
891 % \begin{tabular}{@{}>{\ttfamily}ll@{}}
892 % \toprule
893 % \multicolumn{1}{@{}l}{Value} & Meaning \\
894 % \midrule
895 % AO & Author's Original \\
896 % SMUR & Submitted Manuscript Under Review \\
897 % AM & Accepted Manuscript \\
898 % P & Proof \\
899 % VoR & Version of Record \\
900 % CVoR & Corrected Version of Record \\
901 % EVoR & Enhanced Version of Record \\
902 % \bottomrule
903 % \end{tabular}
904 % \end{table}
905 %
906 % For books, \moptname{pdfbookedition} names the edition of the book.
907 % This is specified as text, not a number. As with
908 % \optname{pdfpublication} (above), \optname{pdfbookedition} accepts a
909 % bracketed language code, as in |pdfbookedition={[en]Second|
910 % |edition}|.
911 %
912 % \acro{XMP} metadata can include a number of dates (in fact,
913 % timestamps, as they include both date and time components).
914 % \moptname{pdfdate} specifies the document date. It is analogous to the
915 % \LaTeX\ |\date| command, and, like |\date|, defaults to the date the
916 % document was built. It must be specified in either \acro{XMP}
917 % format~\cite{Adobe2012:XMP} or \acro{PDF} format~\cite{Adobe2008:PDF}.
918 % \acro{XMP} dates are written in the form
919 % \textsc{yyyy}|-|\textsc{mm}|-|\textsc{dd}|T|hh|:|mm|:|ss|+|\textsc{tt}|:|tt.\footnote{Although
920 % allowed by \acro{XMP}, \pkgname{hyperxmp} does not currently accept
921 % fractions of a second in timestamps.} A W3C
922 % recommendation~\cite{Wolf1997:date-time} discusses this format in more
923 % detail, but as an example, 14~hours, 15~minutes, 9~seconds past
924 % midnight U.S. Mountain Daylight Time (UTC-6) on the 23rd day of
925 % September in the year~2014 should be written as
926 % \texttt{2014-09-23T14:15:09-06:00}. This can be truncated (with loss
927 % of information) to \texttt{2014-09-23T14:15:09},
928 % \texttt{2014-09-23T14:15}, \texttt{2014-09-23}, \texttt{2014-09},
929 % or \texttt{2014} but no other subsets. \acro{PDF} dates are written
930 % in the form |D:|\textsc{yyyymmdd}hhmmss|+|\textsc{tt}|'|tt|'|. The
931 % same date in the preceding example would be written as
932 % \texttt{D:20140923141509-06'00'} in \acro{PDF} format.
933 %
934 % The document's creation date, modification date, and metadata date are
935 % normally set automatically, but \moptname{pdfcreationdate},
936 % \moptname{pdfmoddate}, and \moptname{pdfmetadate} can be used to
937 % override the defaults. Like \optname{pdfdate}, \optname{pdfmetadate}
938 % can be specified in either \acro{XMP} or \acro{PDF} format. However,
939 % because \pkgname{hyperref} defines \optname{pdfcreationdate} and
940 % \optname{pdfmoddate} and expects these to be written as \acro{PDF}
941 % dates, \pkgname{hyperxmp} concomitantly accepts these two dates only
942 % in \acro{PDF} format as well. Note that it's rare that a document
943 % would need to specify any of \optname{pdfcreationdate},
944 % \optname{pdfmoddate}, or \optname{pdfmetadate}.
945 %
946 % \moptname{pdftype} describes the type of document being produced.
947 % This refers to ``the nature or genre of the
948 % resource''~\cite{Adobe2012:XMP} such as |poem|, |novel| or |working|
949 % |paper|, as opposed to the file format (always |application/pdf| when
950 % generated by \pkgname{hyperxmp}). Although \optname{pdftype} can be
951 % assigned an arbitrary piece of text, the \acro{XMP} specification
952 % recommends selecting types from a ``controlled vocabulary'' such as
953 % the \acro{DCMI} Type Vocabulary~\cite{DCMI2012:meta-terms}. The
954 % \acro{DCMI} Type Vocabulary currently consists of only |Collection|,
955 % |Dataset|, |Event|, |Image|, |InteractiveResource|, |MovingImage|,
956 % |PhysicalObject|, |Service|, |Software|, |Sound|, |StillImage|, and
957 % |Text|. \optname{pdftype} defaults to |Text|, which refers to
958 % ``books, letters, dissertations, poems, newspapers, articles, archives
959 % of mailing lists,''~\cite{DCMI2012:meta-terms} and other forms of
960 % text---all things \LaTeX\ is commonly used to typeset.
961 %
962 % Sometimes a base document is rendered in different forms.
963 % \moptname{pdfrendition} indicates the particular rendition the current
964 % document instance represents. The value should come from the
965 % following controlled vocabulary~\cite{Adobe2012:XMP}: |default|,
966 % |draft|, |low-res|, |proof|, |screen|, and |thumbnail|.
967 % \pkgname{hyperxmp}'s default value is |default|, which indicates the
968 % master document, unless the \optname{draft} option is passed to
969 % |\documentclass|, in which case \pkgname{hyperxmp} defaults to
970 % |draft|.
971 %
972 % \pkgname{hyperxmp} honors \pkgname{hyperref}'s \moptname{pdftrapped}
973 % option. A document can indicate whether it employs
974 % \href{https://en.wikipedia.org/wiki/Trap_(printing)}{color trapping}
975 % by specifying \optname{pdftrapped}=|True| or
976 % \optname{pdftrapped}=|False|. (\optname{pdftrapped}=|Unknown| is also
977 % allowed.)
978 %
979 % \moptname{pdfapart} and \moptname{pdfaconformance}, are used in
980 % conjunction with \pkgname{hyperref}'s \optname{pdfa} option to claim a
981 % particular \acro{PDF/A} standard by which the document abides. They
982 % default to \optname{pdfapart}=|1| and \optname{pdfaconformance}=|B|,
983 % indicating the \PDFstd{A}{1}{b}{} standard. These can be changed
984 % (with caution) to assert that the document abides by a different
985 % standard (e.g.,~\PDFstd{A}{2}{u}{}). A document that conforms to the
986 % \acro{PDF/UA} standard can use \moptname{pdfuapart} to indicate the
987 % \acro{PDF/UA} conformance level. For example, \optname{pdfuapart}=|1|
988 % asserts that the document respects \PDFstd{UA}{1}{}{}.
989 % \moptname{pdfxstandard} indicates the particular \acro{PDF/X} standard
990 % by which the document abides. Unlike \optname{pdfapart} and
991 % \optname{pdfaconformance}, which accept a number and a letter,
992 % respectively, \optname{pdfxstandard} expects a textual identification
993 % of a standard name. The following are the acceptable \acro{PDF/X}
994 % standard names as of at the time of this writing.
995 %
996 % ^^A This list needs to be kept up-to-date with the standards checked
997 % ^^A by \@pdfxstandard.
998 % \begin{multicols}{3}
999 % \raggedcolumns
1000 % \begin{itemize}[noitemsep]
1001 % \item \PDFstd{X}{1}{a}{2001}
1002 % \item \PDFstd{X}{1}{a}{2003}
1003 % \item \PDFstd{X}{3}{}{2002}
1004 % \item \PDFstd{X}{3}{}{2003}
1005 % \item \PDFstd{X}{4}{}{}
1006 % \item \PDFstd{X}{4}{p}{}
1007 % \item \PDFstd{X}{5}{g}{}
1008 % \item \PDFstd{X}{5}{n}{}
1009 % \item \PDFstd{X}{5}{pg}{}
1010 % \end{itemize}
1011 % \end{multicols}
1012 %
1013 % For example, one can specify \optname{pdfxstandard}=|{PDF/X-4}| or
1014 % \optname{pdfxstandard}=|{PDF/X-3:2003}|, but specifying
1015 % \optname{pdfxstandard}=|{PDF/X-3}| will not pass \acro{PDF/X}
1016 % validation. Note that at the time of this writing the use of the
1017 % \PDFstd{X}{4}{p}{}, \PDFstd{X}{5}{n}{}, and \PDFstd{X}{5}{pg}{}
1018 % standards has not been tested.
1019 %
1020 % \subsubsection*{Rarely needed options}
1021 %
1022 % \moptname{pdfsource} overrides the name of the \LaTeX\ source file.
1023 % It defaults to |\jobname.tex| but can be replaced by any other string.
1024 % If \optname{pdfsource} is given an empty argument, no document source
1025 % will be specified at all.
1026 %
1027 % The number of pages in the published, print version of the document
1028 % can be expressed with \moptname{pdfnumpages}. This is computed
1029 % automatically when the document is built using either \pdfLaTeX\ or
1030 % \LuaLaTeX\@.
1031 %
1032 % The \moptname{pdfbytes} option expresses the document's file size in
1033 % bytes. The intention is for this to be used to display an estimate of
1034 % download time to a user or to serve as a quick check on whether a file
1035 % was transmitted correctly between systems. \optname{pdfbytes} is
1036 % computed automatically by both \pdfLaTeX\ and \LuaLaTeX, using the
1037 % file size from the previous build of the document.
1038 %
1039 % \bigskip
1040 %
1041 % It is usually more convenient to provide values for all of the options
1042 % presented in this section using \pkgname{hyperref}'s |\hypersetup|
1043 % command than on the |\usepackage| command line. See
1044 % \href{http://mirrors.ctan.org/macros/latex/contrib/hyperref/hyperref.pdf}{the
1045 % \pkgname{hyperref} manual} for more information.
1046 %
1047 %
1048 % \subsection{A complete example}
1049 %
1050 % The following is a sample \LaTeX\ document that provides values for
1051 % most of the metadata options that \pkgname{hyperxmp} recognizes:
1052 %
1053 % \Needspace{4\baselineskip}
1054 % \refstepcounter{pagelabel}
1055 % \label{page:begin-sample-doc}
1056 % \begin{verbatim}
1057 % \documentclass{article}
1058 % \usepackage[utf8]{inputenc}
1059 % \usepackage[unicode]{hyperref}
1060 % \usepackage{hyperxmp}
1061 %
1062 % \title{%
1063 % On a heuristic viewpoint concerning the production and
1064 % transformation of light}
1065 % \author{Albert Einstein}
1066 % \date{March 17, 1905}
1067 %
1068 % \hypersetup{%
1069 % pdftitle={%
1070 % On a heuristic viewpoint concerning the production and
1071 % transformation of light},
1072 % pdfsubtitle={[en-US]Putting that bum Maxwell in his place},
1073 % pdfauthor={Albert Einstein},
1074 % pdfauthortitle={\xmpquote{Technical Assistant\xmpcomma\ Level III}},
1075 % pdfdate={1905-03-17},
1076 % pdfcopyright={Copyright (C) 1905, Albert Einstein},
1077 % pdfsubject={photoelectric effect},
1078 % pdfkeywords={energy quanta, Hertz effect, quantum physics},
1079 % pdflicenseurl={http://creativecommons.org/licenses/by-nc-nd/3.0/},
1080 % pdfcaptionwriter={Scott Pakin},
1081 % pdfcontactaddress={Kramgasse 49},
1082 % pdfcontactcity={Bern},
1083 % pdfcontactpostcode={3011},
1084 % pdfcontactcountry={Switzerland},
1085 % pdfcontactphone={031 312 00 91},
1086 % pdfcontactemail={[email protected]},
1087 % pdfcontacturl={%
1088 % http://einstein.biz/,
1089 % https://www.facebook.com/AlbertEinstein
1090 % },
1091 % pdfdocumentid={uuid:6d1ac9ec-4ff2-515a-954b-648eeb4853b0},
1092 % pdfversionid={2.998e8},
1093 % pdfpublication={[de]Annalen der Physik},
1094 % pdfpublisher={Wiley-VCH},
1095 % pdfpubtype={journal},
1096 % pdfvolumenum={322},
1097 % pdfissuenum={6},
1098 % pdfpagerange={132-148},
1099 % pdfissn={0003-3804},
1100 % pdfeissn={1521-3889},
1101 % pdfpubstatus={VoR},
1102 % pdflang={en},
1103 % pdfmetalang={en},
1104 % pdfurl={http://www.physik.uni-augsburg.de/annalen/history/einstein-papers/1905_17_132-148.pdf},
1105 % pdfdoi={10.1002/andp.19053220607},
1106 % pdfidentifier={info:lccn/50013519}
1107 % }
1108 % \XMPLangAlt{de}{pdftitle={Über einen die Erzeugung und Verwandlung des
1109 % Lichtes betreffenden heuristischen Gesichtspunkt}}
1110 %
1111 % \begin{document}
1112 % \maketitle
1113 % A profound formal difference exists between the theoretical
1114 % concepts that physicists have formed about gases and other
1115 % ponderable bodies, and Maxwell's theory of electromagnetic
1116 % processes in so-called empty space\dots
1117 % \end{document}
1118 % \end{verbatim}
1119 % \refstepcounter{pagelabel}
1120 % \label{page:end-sample-doc}
1121 %
1122 % Compile the document to \acro{PDF} using any of the following
1123 % approaches:
1124 %
1125 % \begin{itemize}
1126 % \item \pdfLaTeX
1127 % \item \LuaLaTeX
1128 % \item \XeLaTeX
1129 % \item \LaTeX~$+$ \term{Dvipdfm}
1130 % \item \LaTeX~$+$ \Dvips~$+$ \term{Ghostscript}
1131 % \item \LaTeX~$+$ \Dvips~$+$ Adobe Acrobat Distiller
1132 % \end{itemize}
1133 %
1134 % \noindent
1135 % The \LaTeX~$+$ \Dvips~$+$ \term{Ghostscript} path stores the
1136 % \acro{XMP} packet in a compressed stream, which implies that a
1137 % \acro{PDF} reader is needed to access it. Ideally, \acro{XMP}
1138 % metadata should be stored uncompressed so it can be extracted as
1139 % ordinary text. Unfortunately, as of 2022-10-07, Ghostscript has no
1140 % plans to support uncompressed metadata streams
1141 % (\href{https://bugs.ghostscript.com/show_bug.cgi?id=705962}{Ghostscript
1142 % bug report~\#705962}). It is possible to leave \emph{all} streams
1143 % uncompressed by passing |-dCompressStreams=false| to \term{Ghostscript}
1144 % (e.g.,~via the \progname{ps2pdf} wrapper script), but this leads to
1145 % larger file sizes.
1146 %
1147 % \bigskip
1148 %
1149 % Once the document is compiled, the resulting \acro{PDF} file will
1150 % contain an \acro{XMP} packet that looks something like that shown in
1151 % Appendix~\ref{sec:sample-packet}. Figure~\ref{fig:xmp-metadata-1} is
1152 % a screenshot of the \acro{XMP} metadata as it appears in Adobe
1153 % Acrobat's ``Advanced'' metadata dialog box. Further clicking on the
1154 % ``Advanced'' item within that dialog box displays all of the
1155 % document's metadata sorted by schema as shown in
1156 % Figure~\ref{fig:xmp-metadata-2}.
1157 %
1158 % \begin{figure}[htbp]
1159 % \includegraphics[width=\linewidth]{einstein1}
1160 % \caption{\acro{XMP} metadata as it appears in Adobe Acrobat}
1161 % \label{fig:xmp-metadata-1}
1162 % \end{figure}
1163 %
1164 % \begin{figure}[htbp]
1165 % \includegraphics[width=\linewidth]{einstein2}
1166 % \caption{Additional \acro{XMP} metadata as it appears in Adobe Acrobat}
1167 % \label{fig:xmp-metadata-2}
1168 % \end{figure}
1169 %
1170 %
1171 % \subsection{Usage notes}
1172 %
1173 % \usagenote{Conflicting metadata in PDF/A documents}
1174 % \label{note:meta-conflicts}
1175 % A \acro{PDF} file includes an \pdfterm{Info} dictionary containing
1176 % \pdfterm{Author}, \pdfterm{Title}, \pdfterm{Subject}, and
1177 % \pdfterm{Keywords} keys. The \pkgname{hyperref} package's
1178 % \optname{pdfauthor}, \optname{pdftitle}, \optname{pdfsubject}, and
1179 % \optname{pdfkeywords} options assign values to those keys. The
1180 % \pkgname{hyperxmp} package additionally uses those options to assign
1181 % values to various \acro{XMP} metadata: \xmpprop{dc:creator},
1182 % \xmpprop{dc:title}, \xmpprop{dc:description}, and
1183 % \xmpprop{pdf:Keywords}. The \acro{PDF/A} specification indicates that
1184 % values that appear in both the \acro{PDF} \pdfterm{Info} dictionary
1185 % and \acro{XMP} packet must match. The problem is that in \acro{XMP},
1186 % the author and keywords can be proper lists, as in
1187 %
1188 % \Needspace{7\baselineskip}
1189 % \begin{verbatim}
1190 % <dc:creator>
1191 % <rdf:Seq>
1192 % <rdf:li>Curly Howard</rdf:li>
1193 % <rdf:li>Larry Fine</rdf:li>
1194 % <rdf:li>Moe Howard</rdf:li>
1195 % </rdf:Seq>
1196 % </dc:creator>
1197 % \end{verbatim}
1198 %
1199 % \noindent
1200 % while in \acro{PDF}, the author and keywords are specified as flat
1201 % strings. Alas, there is no definition of how a list should be
1202 % collapsed to a flat string: ``\texttt{Curly Howard, Larry Fine, Moe
1203 % Howard}'' or ``\texttt{Curly Howard; Larry Fine; Moe Howard}'' or
1204 % something else. I have not yet found a form of flat string that
1205 % passes all \acro{PDF/A} validators. Furthermore, when Adobe
1206 % Acrobat---at least Adobe Acrobat DC (2019) and earlier
1207 % versions---converts a \acro{PDF} file to \acro{PDF/A} format, it does
1208 % so by discarding all but the first author, which is an unsatisfying
1209 % solution.
1210 %
1211 % Starting with version~4.0, \pkgname{hyperxmp}'s solution is to
1212 % suppress writing metadata to the \acro{PDF} \pdfterm{Info} dictionary
1213 % and write it only to the \acro{XMP} packet. (\pkgname{hyperxmp}~v5.0+
1214 % is more sophisticated. It suppresses only the author and keyword
1215 % lists.) This appears to pacify \acro{PDF/A} validators yet retains
1216 % the author and keyword lists in their non-truncated form. If desired,
1217 % the \pdfterm{Info} dictionary can be retained by passing the
1218 % \moptname{keeppdfinfo}\label{page:keeppdfinfo} option to
1219 % \cs{hypersetup}.
1220 %
1221 % \usagenote{Acrobat multiline-field bug}
1222 % The \acro{IPTC} Photo Metadata schema states that ``the [contact]
1223 % address is a multiline field''~\cite{IPTC2010:photo-meta}.
1224 % \pkgname{hyperxmp} converts commas in \optname{pdfcontactaddress}'s
1225 % argument to line breaks in the generated \acro{XML}\@. Unfortunately,
1226 % A bug in Adobe Acrobat---at least in Adobe Acrobat DC (2019) and
1227 % earlier versions---causes that \acro{PDF} reader to discard line
1228 % breaks in the contact address. Interestingly, Adobe Illustrator CS5
1229 % correctly displays the contact address. If you find Adobe Acrobat's
1230 % behavior bothersome, you can redefine the
1231 % \DescribeMacro{\xmplinesep}|\xmplinesep| macro as a string to use as
1232 % an address-line separator. For example, the following replaces all
1233 % commas appearing in \optname{pdfcontactaddress}'s argument with
1234 % semicolons:
1235 %
1236 % \begin{verbatim}
1237 % \renewcommand*{\xmlinesep}{;}
1238 % \end{verbatim}
1239 %
1240 % \usagenote{Object compression}
1241 % One intention of \acro{XMP} is that metadata embedded in a file be
1242 % readable even without knowledge of the file's format. That is, the
1243 % metadata are expected to appear as plain text. Although
1244 % \pkgname{hyperxmp} does its best to honor that intention, it faces
1245 % a few challenges:
1246 %
1247 % \begin{enumerate}
1248 % \item When run with versions of \LuaLaTeX\ earlier than 0.85,
1249 % \pkgname{hyperxmp} leaves all \acro{PDF} objects uncompressed. This
1250 % is due to \LuaLaTeX\ treating object compression as a global
1251 % parameter, unlike \pdfLaTeX, which treats it as a local parameter.
1252 % Hence, when \pkgname{hyperxmp} requests that the \acro{XMP} packet
1253 % be left uncompressed, \LuaLaTeX\ in fact leaves \emph{all}
1254 % \acro{PDF} streams uncompressed. Beginning with version~3.0,
1255 % \pkgname{hyperxmp} includes a workaround that correctly leaves only
1256 % the \acro{XMP} metadata uncompressed, but this workaround is
1257 % implemented only for \LuaLaTeX\ v0.85 onwards.
1258 %
1259 % \item \XeLaTeX\ (or, more precisely, the \cmdname{xdvipdfmx} back end)
1260 % exhibits the opposite problem. It compresses \emph{all} \acro{PDF}
1261 % objects, including the ones containing \acro{XMP} metadata. While
1262 % Adobe Acrobat can still detect and utilize the \acro{XMP} metadata,
1263 % non-\acro{PDF}-aware applications are unlikely to see the metadata.
1264 % Three options to consider are to (1)~use a different program
1265 % (e.g.,~\LuaLaTeX), (2)~pass the |--output-driver="xdvipdfmx -z0"|
1266 % option to \XeLaTeX\ to instruct \cmdname{xdvipdfmx} to turn off all
1267 % compression (which will of course make the \acro{PDF} file
1268 % substantially larger), or (3)~postprocess the generated \acro{PDF}
1269 % file by loading it into the commercial version of Adobe Acrobat and
1270 % re-saving it with the Save As\dots\ menu option.
1271 % \end{enumerate}
1272 %
1273 % \usagenote{Literal commas}
1274 % \label{note:literal-commas}
1275 % \pkgname{hyperxmp} splits the \optname{pdfauthor} and
1276 % \optname{pdfkeywords} lists at commas. Therefore, when specifying
1277 % \optname{pdfauthor} and \optname{pdfkeywords}, you should separate
1278 % items with commas. Also, omit ``|and|'' and other text that does not
1279 % belong to any list item. The following examples should serve as
1280 % clarification:
1281 %
1282 % \begin{description}
1283 % \item[Wrong:] |pdfauthor={Jack Napier, Edward Nigma,|
1284 % \textcolor{red}{\texttt{and}} |Harvey Dent}|
1285 % \item[Wrong:] |pdfauthor={Jack Napier|\textcolor{red}{\texttt{;}}
1286 % |Edward Nigma|\textcolor{red}{\texttt{;}}
1287 % |Harvey Dent}|
1288 % \item[Right:] |pdfauthor={Jack Napier, Edward Nigma, Harvey Dent}|
1289 % \end{description}
1290 %
1291 % \noindent
1292 % \DescribeMacro{\xmpcomma}
1293 % \DescribeMacro{\xmpquote}
1294 % If you need to include a literal comma within an author or keyword
1295 % list (where commas normally separate list items) or a street address
1296 % (where commas normally separate lines), use the |\xmpcomma| macro to
1297 % represent it, and wrap the entire entry containing the comma within
1298 % |\xmpquote{|\dots|}| as shown below:
1299 %
1300 % \begin{verbatim}
1301 % pdfauthor={\xmpquote{Jack Napier\xmpcomma\ Jr.},
1302 % \xmpquote{Edward Nigma\xmpcomma\ PhD},
1303 % \xmpquote{Harvey Dent\xmpcomma\ Esq.}}
1304 %
1305 % pdfcontactaddress={Office of the President,
1306 % \xmpquote{Wayne Enterprises\xmpcomma\ Inc.},
1307 % One Wayne Blvd}
1308 % \end{verbatim}
1309 %
1310 % As of version~2.2 of \pkgname{hyperxmp}, it is acceptable to use
1311 % |\xmpcomma| and |\xmpquote| within any \pkgname{hyperxmp} option, not
1312 % just in those in which a comma normally serves as a separator
1313 % (i.e.,~lists and multiline fields). Outside of cases in which a comma
1314 % serves as a separator, |\xmpcomma| is treated as an ordinary comma,
1315 % and |\xmpquote| returns its argument unmodified. Hence, it is
1316 % legitimate to use |\xmpcomma| and |\xmpquote| in cases like the
1317 % following
1318 %
1319 % \begin{verbatim}
1320 % pdfauthortitle={\xmpquote{Psychiatrist\xmpcomma\ Arkham Asylum}}
1321 % \end{verbatim}
1322 %
1323 % \noindent
1324 % (Like most \pkgname{hyperxmp} options, \optname{pdfauthortitle}
1325 % inserts its argument unmodified in an \acro{XMP} tag.) When in
1326 % doubt, use |\xmpcomma| and |\xmpquote|; it should always be safe
1327 % to do so.
1328 %
1329 % \DescribeMacro{\xmptilde}
1330 % Version~2.4 of \pkgname{hyperxmp} introduces a convenience macro
1331 % called |\xmptilde|. |\xmptilde| expands to a literal tilde character
1332 % instead of the nonbreaking space that ``|~|'' normally represents.
1333 % Use it to represent \acro{URL}s such as
1334 % \url{http://www.pakin.org/~scott/} (``|http://www.pakin.org/|\linebreak[0]|\xmptilde|
1335 % |scott/|'') in options such as \optname{baseurl},
1336 % \optname{pdfcontacturl} and \optname{pdflicenseurl}.
1337 %
1338 % \usagenote{Unicode support}
1339 % \term{Unicode} support is provided via the \pkgname{hyperref} package.
1340 % If you specify \optname{unicode}|=true| either as a \pkgname{hyperref}
1341 % option or as an argument to the |\hypersetup| command, the document
1342 % can include \term{Unicode} characters in its \acro{XMP} fields.
1343 %
1344 % \usagenote{Automatically specified metadata}
1345 % \pkgname{hyperxmp} attempts to identify certain metadata
1346 % automatically. The hope is that in many cases, an author can simply
1347 % include |\usepackage{hyperxmp}| in a document's preamble and benefit
1348 % from a modicum of \acro{XMP} metadata with no additional effort.
1349 %
1350 % Currently, \optname{pdftitle} defaults to the document's title as
1351 % specified by |\title{|\dots|}|. \optname{pdfauthor} defaults to the
1352 % document's author(s) as specified by |\author{|\dots|}|.
1353 % \optname{pdfdate} defaults to the current date and time.
1354 % \optname{pdfmetalang} defaults to the same value as \optname{pdflang}
1355 % if non-empty, ``\xmpterm{x-default}'' otherwise. \pkgname{hyperxmp}
1356 % recognizes some class-specific metadata as well, such as that provided
1357 % via the \Koma\ letter classes (e.g.,~\clsname{scrlttr2}) and the
1358 % \ACMclass\ article class (\clsname{acmart}).
1359 %
1360 % If a document uses either the \pkgname{babel} or \pkgname{polyglossia}
1361 % packages it is recommended that it \emph{not} explicitly set
1362 % \optname{pdflang}. \optname{pdflang} accepts only a single language
1363 % name while \pkgname{hyperxmp} can automatically query \pkgname{babel}
1364 % and \pkgname{polyglossia} for a list of all languages used in the
1365 % document and include this list in an \acro{XMP} \xmpprop{dc:language}
1366 % element.
1367 %
1368 % \usagenote{Multilingual metadata}
1369 % \label{note:multilingual}
1370 % The \optname{pdfmetalang} option specifies the language in which the
1371 % document's metadata is written. It defaults to the value of
1372 % \optname{pdflang}, which specifies the document language.
1373 % \DescribeMacro{\XMPLangAlt} As of version~3.3 of \pkgname{hyperxmp},
1374 % it is possible to include certain metadata---specifically, the
1375 % document's title, subject, and copyright statement---in more than one
1376 % language. The |\XMPLangAlt| macro provides this functionality. Usage
1377 % is as follows:
1378 %
1379 % \bigskip
1380 % \indent
1381 % |\XMPLangAlt| \marg{language} |{| \meta{option}|=|\meta{text},~\dots\ |}|
1382 % \bigskip
1383 %
1384 % \noindent
1385 % where \meta{language} is an \acro{ISO} \mbox{639-1} two-letter country
1386 % code with an optional \acro{ISO} \mbox{3166-1} two-letter region code
1387 % (e.g.,~``|en|'' for English or ``|en-US|'' for specifically US
1388 % English); \meta{option} is one of ``|pdftitle|'', ``|pdfsubject|'', or
1389 % ``|pdfcopyright|''; and \meta{text} is the text as expressed in the
1390 % specified language. By way, of example, the following code provides
1391 % the document title in English then specifies an alternative title to
1392 % use in four other languages:
1393 %
1394 % \begin{verbatim}
1395 % \hypersetup{%
1396 % pdfmetalang={en},
1397 % pdftitle={English title}
1398 % }
1399 % \XMPLangAlt{de}{pdftitle={Deutscher Titel}}
1400 % \XMPLangAlt{fr}{pdftitle={Titre fran\c{c}ais}}
1401 % \XMPLangAlt{it}{pdftitle={Titolo italiano}}
1402 % \XMPLangAlt{rm}{pdftitle={Titel rumantsch}}
1403 % \end{verbatim}
1404 %
1405 % \usagenote{Expandable arguments}
1406 % \label{note:expandable}
1407 % All arguments passed to \pkgname{hyperxmp} options must be expandable,
1408 % in \tex\ terminology. This implies that any macros that are used in
1409 % arguments are limited to a relatively small set of operations (such as
1410 % conditionals and macro expansion) and must produce a string of text.
1411 % Code (such as macro definitions and arithmetic operations) will be
1412 % written to \acro{XMP} as code, not as the result of executing the
1413 % code.
1414 %
1415 % By way of example, the macros provided by the \pkgname{texdate}
1416 % package for typesetting dates are not expandable (at least at the time
1417 % of this writing). Hence, the |\printfdate{Y}| in the following code
1418 % snippet is not replaced by the current year, as one might expect:
1419 %
1420 % \begin{verbatim}
1421 % \usepackage{texdate}
1422 % \initcurrdate
1423 % \hypersetup{%
1424 % pdfcopyright={Copyright \textcopyright\ \printfdate{Y}, Scott Pakin}
1425 % }
1426 % \end{verbatim}
1427 %
1428 % \noindent
1429 % Rather, it generates a \xmpprop{dc:rights} tag of the form
1430 % ``\texttt{Copyright \textcopyright\ =2=0=by-1by=0\the\year, Scott
1431 % Pakin}''. The garbage in that line corresponds to the remnants of
1432 % the |\printfdate| code after expanding all of the \tex\ primitives and
1433 % certain other control sequences it uses to the empty string. For
1434 % example, ``|\global\advance\texd@yr| |by-1|'' expands to ``|by-1|''.
1435 %
1436 % It is not possible to determine a~priori whether or not a macro is
1437 % expandable. The best advice is to carefully inspect the \acro{XMP}
1438 % package in the output file to ensure that any macros used in arguments
1439 % to \pkgname{hyperxmp} options produced the expected output.
1440 %
1441 % \usagenote{Semi-automatic page ranges}
1442 % \label{note:page-counting}
1443 % Although \optname{pdfpagerange} is intended to refer to pages in the
1444 % final, published version of a document, it would be convenient for
1445 % them to be generated automatically when producing a standalone
1446 % \acro{PDF} file that is not intended to be incorporated into a book,
1447 % journal, or other publication (or if it is known that the pages will
1448 % not be renumbered for publication). One approach is to use the
1449 % \pkgname{totpages} package help generate \optname{pdfpagerange}. For
1450 % documents numbered from~1 to~$n$, a simple
1451 %
1452 % \Needspace{3\baselineskip}
1453 % \begin{verbatim}
1454 % \hypersetup{%
1455 % pdfpagerange={1-\ref*{TotPages}}
1456 % }
1457 % \end{verbatim}
1458 %
1459 % \noindent
1460 % should suffice. A bit more effort is needed for documents that change
1461 % numbering schemes, such as using lowercase Roman numerals for the
1462 % front matter and Arabic numerals for the main matter and back matter.
1463 % One approach is to use \cs{label} to mark the first and last page of
1464 % each numbering scheme and specify \optname{pdfpagerange} as in the
1465 % following:
1466 %
1467 % \Needspace{6\baselineskip}
1468 % \begin{verbatim}
1469 % \hypersetup{%
1470 % pdfpagerange={%
1471 % \pageref*{page:begin-front}-\pageref*{page:end-front},%
1472 % 1-\pageref*{TotPages}%
1473 % }
1474 % }
1475 % \end{verbatim}
1476 %
1477 % I don't know how unnumbered pages (e.g.,~blank pages and the title
1478 % page) are supposed to be handled. I suppose blank pages can be
1479 % omitted from \optname{pdfpagerange}, and the title page can be either
1480 % omitted or listed as \texttt{title}, for example.
1481 %
1482 % It appears that at least with version~2.00 of \pkgname{totpages}, the
1483 % |TotPages| label is not defined until after the |\begin{document}|.
1484 % Consequently, using |TotPages| within a \cs{hypersetup} invocation in
1485 % the document's preamble will produce ``|??|'' as the page count in the
1486 % \acro{XMP} packet. The solution is either to assign
1487 % \optname{pdfpagerange} after the |\begin{document}| or to ask
1488 % \LaTeX\ to do that on your behalf:
1489 %
1490 % \begin{verbatim}
1491 % \AtBeginDocument{%
1492 % \hypersetup{%
1493 % pdfpagerange={1-\ref*{TotPages}}
1494 % }%
1495 % }
1496 % \end{verbatim}
1497 %
1498 % \usagenote{Automatic computation of the PDF byte count}
1499 % \label{note:auto-byte-count}
1500 %
1501 % The \schema[PRISM Basic Metadata]{\protect\acro{PRISM} Basic Metadata}
1502 % schema~\cite{PRISM2012:basic-meta} defines a \xmpprop{prism:byteCount}
1503 % property that indicates the \acro{PDF} file size in bytes.
1504 % \pkgname{hyperxmp} computes this value automatically when the document
1505 % is built using \LuaLaTeX\ but not when using any other \TeX\ engine.
1506 % Note that \pkgname{hyperxmp} uses the file size from the
1507 % \emph{previous} run of \LuaLaTeX\ because the new \acro{PDF} file is
1508 % not yet complete. Consequently, one extra compilation is needed for
1509 % the byte count to converge relative to the the number of compilations
1510 % that would otherwise be required.
1511 %
1512 % Starting with \pkgname{hyperxmp}~v5.9, the \pkgname{hyperxmp}
1513 % distribution includes a Perl\index{Perl} script called
1514 % \progname{hyperxmp-add-bytecount} that edits a \acro{PDF} file in
1515 % place, adding or replacing the \xmpprop{prism:byteCount} property with
1516 % one that specifies the final file size.\footnote{The script was in
1517 % fact introduced with \progname{hyperxmp}~v5.8 and was then called
1518 % \texttt{add\_byteCount}.} Run the script as
1519 % ``\progname{hyperxmp-add-bytecount} \meta{filename.pdf}''.
1520 %
1521 % The \progname{latexmk} build tool can be configured to run
1522 % \progname{hyperxmp-add-bytecount} automatically every time a \acro{PDF} file
1523 % is generated. Simply add the code shown in Figure~\ref{fig:latexmkrc}
1524 % to your \progname{latexmk} configuration file. See
1525 % \href{http://mirrors.ctan.org/support/latexmk/latexmk.pdf}{the
1526 % \progname{latexmk} manual} for information on configuration-file
1527 % naming on different operating systems and explanations of the hook
1528 % functions used in Figure~\ref{fig:latexmkrc}.
1529 %
1530 % \changes{v5.8}{2020/11/18}{Distribute an
1531 % \protect\progname{add\_byteCount} script and document some sample
1532 % \protect\progname{latexmk} configuration code that invokes it.
1533 % Thanks to John Collins for providing both of those}
1534 % \changes{v5.9}{2020/11/22}{At Karl Berry's request, rename
1535 % \protect\progname{add\_byteCount} to the less generic-sounding
1536 % \protect\progname{hyperxmp-add-bytecount}}
1537 %
1538 % \iffalse
1539 %<*listings>
1540 % \fi
1541 \newsavebox{\latexmkrcbox}
1542 \begin{lrbox}{\latexmkrcbox}
1543 \begin{lstlisting}[
1544 language=perl,
1545 columns=fullflexible,
1546 linewidth=19pc,
1547 frame=single,
1548 basicstyle=\footnotesize,
1549 showstringspaces=false,
1550 upquote=true,
1551 literate={-}{-}{1}
1552 ]
1553 foreach my $cmd ( "latex", "lualatex", "pdflatex", "xelatex",
1554 "dvipdf", "xdvipdfmx", "ps2pdf" ) {
1555 ${$cmd} = "internal mycmd ${$cmd}";
1556 }
1557
1558 sub mycmd {
1559 my $retval = system @_;
1560 if ( $$Pdest =~ /\.pdf$/ ) {
1561 system 'hyperxmp-add-bytecount', $$Pdest;
1562 }
1563 return $retval;
1564 }
1565 \end{lstlisting}
1566 \end{lrbox}
1567 % \iffalse
1568 %</listings>
1569 % \fi
1570 %
1571 % \begin{figure}
1572 % \centering
1573 % \usebox{\latexmkrcbox}
1574 % \caption{ \progname{latexmk} configuration-file code for
1575 % automatically invoking \progname{hyperxmp-add-bytecount} every time a
1576 % \acro{PDF} file is generated}
1577 % \label{fig:latexmkrc}
1578 % \end{figure}
1579 %
1580 % Even though \pkgname{hyperxmp} can compute the byte count
1581 % automatically when run from \LuaLaTeX, users of \progname{latexmk}
1582 % need to use configuration-file code like that shown in
1583 % Figure~\ref{fig:latexmkrc}. Otherwise, \progname{latexmk} would
1584 % compile the document one time too few for the byte count to
1585 % converge. It is recommended that those who use both \progname{latexmk}
1586 % and \pkgname{hyperxmp} configure \progname{latexmk} to be
1587 % \pkgname{hyperxmp}-aware.
1588 %
1589 %
1590 % \StopEventually{^^A
1591 % \section{Help Wanted}
1592 %
1593 % \paragraph{Comma handling}
1594 % Ideally, \texttt{\string\xmpquote} should automatically replace all
1595 % commas with \texttt{\string\xmpcomma}. Unfortunately, my \tex\ skills
1596 % are insufficient to pull that off. If you know a way to make
1597 % \texttt{\string\xmpquote\string{Hello, world\string}} work with both
1598 % \term{Unicode} and non-\term{Unicode} encodings and with all
1599 % \tex\ engines (\pdfTeX, \LuaTeX, \XeTeX, etc.), please send me a code
1600 % patch.
1601 %
1602 % \appendix
1603 %
1604 % \section{Sample XMP Packet}
1605 % \label{sec:sample-packet}
1606 %
1607 % The following is an example of a complete \acro{XMP} packet as may be
1608 % produced by \pkgname{hyperxmp}. This packet corresponds to the
1609 % metadata included in the sample \LaTeX\ document presented
1610 % \vpagerefrange{page:begin-sample-doc}{page:end-sample-doc}. For
1611 % clarity, metadata values, either specified explicitly by the document
1612 % or introduced automatically by \pkgname{hyperxmp}, are colored blue.
1613 %
1614 % \input{einstein-xmp}
1615 %
1616 % \begin{thebibliography}{99}
1617 % \bibitem{NISO2008:jav}
1618 % Beverley Acreman, Claire Bird, Catherine Jones, Peter McCracken, Cliff Morgan,
1619 % John Ober, Evan Owens, T.~Scott Plutchak, Bernie Rous, and Andrew Wray.
1620 % \newblock {J}ournal {A}rticle {V}ersions ({JAV}): Recommendations of the
1621 % {NISO}/{ALPSP} {JAV} {T}echnical {W}orking {G}roup.
1622 % \newblock Recommended practice, National Information Standards Organization,
1623 % Baltimore, Maryland, USA, April 2008.
1624 % \newblock ISBN~\mbox{978-1-880124-79-6}.
1625 % \newblock Available from
1626 % \url{https://www.niso.org/sites/default/files/2017-08/RP-8-2008.pdf}.
1627 %
1628 % \bibitem{Adobe2010:pdfmark}
1629 % Adobe Systems, Inc., San Jose, California.
1630 % \newblock \emph{{A}dobe {A}crobat~{X} {SDK} Help, pdfmark Reference}.
1631 % \newblock Available from
1632 % \url{http://www.adobe.com/devnet/acrobat/documentation.html}.
1633 %
1634 % \bibitem{Adobe1996:postscript}
1635 % {Adobe Systems, Inc.}
1636 % \newblock \emph{{P}ost{S}cript Language Reference Manual}.
1637 % \newblock Addison-Wesley, 2nd edition, January 1996, ISBN: \mbox{0-201-18127-4}.
1638 %
1639 % \bibitem{Adobe2008:PDF}
1640 % Adobe Systems, Inc., San Jose, California.
1641 % \newblock \emph{Document Management---Portable Document Format---Part 1:
1642 % PDF~1.7}, July 2008.
1643 % \newblock ISO \mbox{32000-1} standard document. Available from
1644 % \url{http://wwwimages.adobe.com/www.adobe.com/content/dam/Adobe/en/devnet/pdf/pdfs/PDF32000_2008.pdf}.
1645 %
1646 % \bibitem{Adobe2012:XMP}
1647 % Adobe Systems, Inc., San Jose, California.
1648 % \newblock \emph{{XMP} Specification Part~1: Data model, Serialization, and Core
1649 % Properties}, April 2012.
1650 % \newblock Available from
1651 % \url{http://wwwimages.adobe.com/www.adobe.com/content/dam/Adobe/en/devnet/xmp/pdfs/cc-201306/XMPSpecificationPart1.pdf}.
1652 %
1653 % \bibitem{DCMI2012:meta-terms}
1654 % DCMI Usage Board
1655 % \newblock \emph{DCMI Metadata Terms}, June~14, 2012.
1656 % \newblock Available from \url{http://dublincore.org/documents/dcmi-terms/}.
1657 %
1658 % \bibitem{Downes1994:ATB15}
1659 % Michael Downes.
1660 % \newblock Around the bend~\#15, answers, 4th (last) installment.
1661 % \newblock \href{news:comp.text.tex}{\texttt{comp.text.tex}} newsgroup posting,
1662 % January~3, 1994.
1663 % \newblock Archived by Google at
1664 % \url{http://groups.google.com/group/comp.text.tex/msg/7da7643b9e8f3b48}.
1665 %
1666 % \bibitem{PRISM2012:basic-meta}
1667 % International Digital Enterprise Alliance, Inc.
1668 % \newblock \emph{Publishing Requirements for Industry Standard Metadata,
1669 % Version~3.0: {PRISM} Basic Metadata Specification}, October~12, 2012.
1670 % \newblock Available from
1671 % \url{http://www.prismstandard.org/specifications/3.0/PRISM_Basic_Metadata_3.0.htm}.
1672 %
1673 % \bibitem{PRISM2012:cont-voc}
1674 % International Digital Enterprise Alliance, Inc.
1675 % \newblock \emph{Publishing Requirements for Industry Standard Metadata,
1676 % Version~3.0: {PRISM} Controlled Vocabularies Specification}, October~4, 2012.
1677 % \newblock Available from
1678 % \url{http://www.prismstandard.org/specifications/3.0/PRISM_CV_Spec_3.0.pdf}.
1679 %
1680 % \bibitem{IPTC2010:photo-meta}
1681 % International Press Telecommunications Council.
1682 % \newblock \emph{{IPTC} Photo Metadata: Core 1.1/Extension 1.1}, July 2010.
1683 % \newblock Revision~1. Available from
1684 % \url{http://www.iptc.org/std/photometadata/specification/IPTC-PhotoMetadata-201007_1.pdf}.
1685 %
1686 % \bibitem{IANA2011:lang-tags}
1687 % {Internet Assigned Numbers Authority}.
1688 % \newblock Language subtag registry, January~11, 2011.
1689 % \newblock Available from
1690 % \url{http://www.iana.org/assignments/language-subtag-registry}.
1691 %
1692 % \bibitem{Leach2005:uuid}
1693 % Paul~J. Leach, Michael Mealling, and Rich Salz.
1694 % \newblock A {U}niversally {U}nique {ID}entifier ({UUID}) {URN} namespace.
1695 % \newblock Request for Comments 4122, Internet Engineering Task Force, Network
1696 % Working Group, July 2005.
1697 % \newblock Category: Standards Track. Available from
1698 % \url{http://www.ietf.org/rfc/rfc4122.txt}.
1699 %
1700 % \bibitem{PDFA2008:xmp-props}
1701 % PDF/A Competence Center, Berlin, Germany.
1702 % \newblock \emph{{T}ech{N}ote~0008: Predefined {XMP} Properties in {PDF/A-1}},
1703 % March~20, 2008.
1704 % \newblock Available from
1705 % \url{http://www.pdfa.org/wp-content/uploads/2011/08/tn0008_predefined_xmp_properties_in_pdfa-1_2008-03-20.pdf}.
1706 %
1707 % \bibitem{PDFA2008:ext-schemas}
1708 % PDF/A Competence Center, Berlin, Germany.
1709 % \newblock \emph{{T}ech{N}ote~0009: {XMP} Extension Schemas in {PDF/A-1}},
1710 % March~20, 2008.
1711 % \newblock Available from
1712 % \url{http://www.pdfa.org/wp-content/uploads/2011/08/tn0009_xmp_extension_schemas_in_pdfa-1_2008-03-20.pdf}.
1713 %
1714 % \bibitem{Wolf1997:date-time}
1715 % Misha Wolf and Charles Wicksteed.
1716 % \newblock Date and time formats.
1717 % \newblock Note NOTE-datetime, World Wide Web Consortium (W3C), September~15,
1718 % 1997.
1719 % \newblock Available from \url{http://www.w3.org/TR/NOTE-datetime}.
1720 % \end{thebibliography}
1721 % }
1722 %
1723 %
1724 % \section{Implementation}
1725 %
1726 % This section presents the commented \LaTeX\ source code for
1727 % \pkgname{hyperxmp}. Read this section only if you want to learn how
1728 % \pkgname{hyperxmp} is implemented.
1729 %
1730 % \bigskip
1731 %
1732 % One thing to bear in mind when reading the \pkgname{hyperxmp} source
1733 % code is that different actions occur at different times throughout
1734 % document processing:
1735 %
1736 % \begin{enumerate}
1737 % \item |\usepackage{hyperxmp}|: \pkgname{hyperxmp} parses package
1738 % options, defines a number of commands, loads various helper
1739 % packages, and assigns default values to most \acro{XMP} fields.
1740 %
1741 % \item |\begin{document}|: \pkgname{hyperxmp} loads certain packages
1742 % such as \pkgname{hyperref} and \pkgname{ifdraft} and queries
1743 % natural-language information from \pkgname{babel} and
1744 % \pkgname{polyglossia} that becomes available only at the end of
1745 % the preamble.
1746 %
1747 % \item |\end{document}|: \pkgname{hyperxmp} finalizes certain data
1748 % that are known only at the end of the document, such as the page
1749 % count, and writes the \acro{XMP} packet to the \acro{PDF} file.
1750 % \end{enumerate}
1751 %
1752 % \iffalse
1753 %<*package>
1754 % \fi
1755 %
1756 % \subsection{Initial preparation}
1757 % \label{sec:initial-prep}
1758 %
1759 % \changes{v5.11}{2022/10/02}{Disable \protect\pkgname{hyperxmp} if
1760 % \LaTeX3 document metadata is available. Document metadata implies
1761 % the presence of PDF management, which completely breaks
1762 % \protect\pkgname{hyperxmp}}
1763 % \begin{macrocode}
1764 \IfDocumentMetadataTF{%
1765 \PackageWarning
1766 {hyperxmp}
1767 {Disabling hyperxmp because it is incompatible with PDF management}
1768 }{}
1769 \IfDocumentMetadataTF{\endinput}{}
1770 % \end{macrocode}
1771 %
1772 % \changes{v1.2}{2011/04/17}{Made the package compatible with \pkgname{ngerman}.
1773 % Thanks to Tobias Mueller for the bug report.}
1774 %
1775 % \begin{macro}{\hyxmp@dq@code}
1776 % The \pkgname{ngerman} package redefines
1777 % ``\,{\fontencoding{T1}\selectfont\textquotedbl}\,'' as an active
1778 % character, which causes problems for \pkgname{hyperxmp} when it tries
1779 % to use that character. We therefore save the double-quote character's
1780 % current category code in |\hyxmp@dq@code| and mark the character as
1781 % category code~12 (``other''). The original category code is restored
1782 % at the end of the package code (Section~\ref{sec:clean-up}).
1783 % \begin{macrocode}
1784 \edef\hyxmp@dq@code{\the\catcode`\"}
1785 \catcode`\"=12
1786 % \end{macrocode}
1787 % \end{macro}
1788 %
1789 % \changes{v1.5}{2012/03/10}{Made the \protect\acro{XMP} inclusion more robust.
1790 % Thanks to Heiko Oberdiek for the bug report and suggested modifications.}
1791 %
1792 % \begin{macro}{\hyxmp@at@end}
1793 % The |\hyxmp@at@end| macro includes code at the end of the document.
1794 % When available (as is the case in most modern \tex\ backends),
1795 % \cs{AtEndDocument} works well enough. Otherwise, we invoke
1796 % \cs{AtEndDvi} from the \pkgname{atenddvi} package, which is robust but
1797 % requires an addition \LaTeX\ run.
1798 % \changes{v5.3}{2020/05/29}{Use \protect\cs{AtEndDocument} in all
1799 % \protect\tex\ back ends that provide it. Thanks to Nelson Posse
1800 % Lago for pointing out why \protect\pkgname{atenddvi} is best avoided
1801 % if possible}
1802 % \changes{v5.11}{2022/10/16}{Use \protect\cs{AddToHook} when available.
1803 % This addresses a bug reported on \protect\TeX\ StackExchange by joHub
1804 % and solved by Ulrike Fischer}
1805 % \begin{macrocode}
1806 \@ifundefined{AddToHook}{%
1807 \@ifundefined{AtEndDocument}{%
1808 \RequirePackage{atenddvi}
1809 \let\hyxmp@at@end=\AtEndDvi
1810 }{%
1811 \let\hyxmp@at@end=\AtEndDocument
1812 }
1813 }{%
1814 \def\hyxmp@at@end{\AddToHook{shipout/lastpage}}
1815 }
1816 % \end{macrocode}
1817 % \end{macro}
1818 %
1819 % \begin{macro}{\hyxmp@set@jobname}
1820 % Given an expanded \cs{jobname} followed by \cs{relax}, invoke the
1821 % \cs{hyxmp@set@jobname@dbl} macro if the job name is surrounded by
1822 % double quotes and the \cs{hyxmp@set@jobname@plain} macro otherwise.
1823 % \begin{macrocode}
1824 \def\hyxmp@set@jobname#1\relax{%
1825 \@ifnextchar"{\hyxmp@set@jobname@dbl}{\hyxmp@set@jobname@plain}#1\relax
1826 }
1827 % \end{macrocode}
1828 % \end{macro}
1829 %
1830 % \begin{macro}{\hyxmp@set@jobname@dbl}
1831 % \begin{macro}{\hyxmp@jobname}
1832 % Set \cs{hyxmp@jobname} to to |#1|, discarding the surrounding double
1833 % quotes.
1834 % \begin{macrocode}
1835 \def\hyxmp@set@jobname@dbl"#1"\relax{\xdef\hyxmp@jobname{#1}}
1836 % \end{macrocode}
1837 % \end{macro}
1838 % \end{macro}
1839 %
1840 % \begin{macro}{\hyxmp@set@jobname@plain}
1841 % \begin{macro}{\hyxmp@jobname}
1842 % Set \cs{hyxmp@jobname} to to |#1|.
1843 % \begin{macrocode}
1844 \def\hyxmp@set@jobname@plain#1\relax{\xdef\hyxmp@jobname{#1}}
1845 % \end{macrocode}
1846 % \end{macro}
1847 % \end{macro}
1848 %
1849 % Define \cs{hyxmp@jobname} as a sanitized version of \cs{jobname}. The
1850 % problem with using \cs{jobname} directly is that it surrounds the
1851 % filename with double quotes if it contains a space character. For
1852 % example, a source file named \texttt{my-file.tex} results in a
1853 % \cs{jobname} of ``\texttt{my-file}'', but a source file named
1854 % \texttt{my file.tex} results in a \cs{jobname} of ``|"my file"|''.
1855 % Trying to access |"my file".log| (as is done on
1856 % page~\pageref{page:jobname-log}) will fail because the filename does
1857 % not in fact contain literal double quotes.
1858 % \changes{v5.5}{2020/07/28}{Correctly handle source files with spaces
1859 % in their name. Thanks to Peter Dyballa for the bug report}
1860 % \begin{macrocode}
1861 \expandafter\hyxmp@set@jobname\jobname\relax
1862 % \end{macrocode}
1863 %
1864 % \begin{macro}{\hyxmp@aep@toks}
1865 % In order for \pkgname{hyperxmp} to be loaded safely during
1866 % \cs{AtEndPreamble} we need to ensure that we perform no
1867 % \cs{AtEndPreamble} actions until all top-level macro definitions have
1868 % been made. The most straightforward approach would be to move all of
1869 % \pkgname{hyperxmp}'s \cs{AtEndPreamble} stanzas to the end of the
1870 % package. However, this degrades readability of the source code. For
1871 % instance, an \cs{AtEndPreamble} stanza related to integration with
1872 % \pkgname{hyperref} could no longer appear in the ``Integration with
1873 % \pkgname{hyperref}'' section (Section~\ref{sec:hyperref-int}). Hence,
1874 % we instead store in a token list, \cs{hyxmp@aep@toks}, each
1875 % \cs{AtEndPreamble} stanza as we encounter it. This token list is
1876 % evaluated as one of the package's final actions
1877 % (Section~\ref{sec:clean-up}).
1878 % \begin{macrocode}
1879 \newtoks{\hyxmp@aep@toks}
1880 % \end{macrocode}
1881 % \end{macro}
1882 %
1883 %
1884 % \subsection{Integration with \textsf{hyperref}}
1885 % \label{sec:hyperref-int}
1886 %
1887 % An important design decision underlying \pkgname{hyperxmp} is that the
1888 % package should integrate seamlessly with \pkgname{hyperref}. To that
1889 % end, \pkgname{hyperxmp} takes \acro{XMP} metadata from
1890 % \pkgname{hyperref}'s \optname{baseurl}, \optname{pdfauthor},
1891 % \optname{pdfkeywords}, \optname{pdflang}, \optname{pdfproducer},
1892 % \optname{pdfsubject}, \optname{pdftrapped}, and \optname{pdftitle}
1893 % options. It also introduces a number of new options, which are listed
1894 % \vpagerefrange[above]{page:begin-new-options}{page:end-new-options}.
1895 % For consistency with \pkgname{hyperref}'s document-metadata naming
1896 % conventions (which are in turn based on \LaTeX's document-metadata
1897 % naming conventions), we do not prefix metadata-related macro names
1898 % with our package-specific |\hyxmp@| prefix. That is, we use names
1899 % like |\@pdfcopyright| instead of |\hyxmp@pdfcopyright|.
1900 %
1901 % \bigskip
1902 %
1903 % We load a bunch of helper packages: \pkgname{kvoptions} for
1904 % package-option processing, \pkgname{pdfescape} and \pkgname{stringenc}
1905 % for re-encoding \term{Unicode} strings, \pkgname{intcalc} for
1906 % performing integer calculations (division and modulo), \pkgname{iftex}
1907 % for determining which \tex\ engine is being used, \pkgname{ifmtarg}
1908 % for testing if a macro argument is empty or all spaces,
1909 % \pkgname{etoolbox} for dynamically patching existing commands
1910 % (specifically, \pkgname{hyperref}'s |\PDF@FinishDoc|), and
1911 % \pkgname{ifthen} for convenient string comparisons.
1912 % \begin{macrocode}
1913 \RequirePackage{kvoptions}
1914 \RequirePackage{pdfescape}
1915 \RequirePackage{stringenc}
1916 \RequirePackage{intcalc}
1917 \RequirePackage{iftex}
1918 \RequirePackage{ifmtarg}
1919 \RequirePackage{etoolbox}
1920 \RequirePackage{ifthen}
1921 % \end{macrocode}
1922 %
1923 % There are a few places where \pkgname{hyperxmp} can take advantage of
1924 % \LuaTeX\ features. To simplify the use of \LuaTeX\ we load the
1925 % \pkgname{luacode} package.
1926 % \begin{macrocode}
1927 \ifLuaTeX
1928 \RequirePackage{luacode}
1929 \fi
1930 % \end{macrocode}
1931 %
1932 % \begin{macro}{\@ifmtargexp}
1933 % \begin{macro}{\@ifnotmtargexp}
1934 % |\@ifmtarg| and |\@ifnotmtarg| do not expand their first argument.
1935 % Define |\@ifmtargexp| and |\@ifnotmtargexp| as expanding versions of
1936 % those macros.
1937 % \begin{macrocode}
1938 \def\@ifmtargexp#1{\expandafter\@ifmtarg\expandafter{#1}}
1939 \def\@ifnotmtargexp#1{\expandafter\@ifnotmtarg\expandafter{#1}}
1940 % \end{macrocode}
1941 % \end{macro}
1942 % \end{macro}
1943 %
1944 % \begin{macro}{\@if@def@and@nonempty}
1945 % This macro combines \cs{@ifundefined} and \cs{@ifmtargexp}. If the
1946 % macro named |#1| is both defined and non-empty, evaluate~|#2|.
1947 % Otherwise, evaluate~|#3|.
1948 % \changes{v5.3}{2020/06/07}{Added this macro}
1949 % \begin{macrocode}
1950 \newcommand*{\@if@def@and@nonempty}[3]{%
1951 \@ifundefined{#1}{#3}{%
1952 \expandafter\@ifmtargexp\expandafter{\csname#1\endcsname}{#3}{#2}%
1953 }%
1954 }
1955 % \end{macrocode}
1956 % \end{macro}
1957 %
1958 % \begin{macro}{\hyxmp@pdfstringdef}
1959 % \begin{macro}{\hyxmp@textunderscore}
1960 % Because \pkgname{hyperxmp} uses underscores to represent hard spaces,
1961 % we need ``|\_|'' to map initially to something other than an
1962 % underscore, in particular the \acro{ASCII} \acro{NAK}~(|^^U|)
1963 % character. To accomplish this, we wrap \pkgname{hyperref}'s
1964 % |\pdfstringdef| macro with our own version that temporarily does the
1965 % proper substitution. Later in the execution, after underscores have
1966 % been replaced with spaces, we replace \acro{NAK} characters with
1967 % underscores.
1968 % \changes{v2.5}{2014/06/19}{Added this macro}
1969 % \begin{macrocode}
1970 \newcommand{\hyxmp@pdfstringdef}[2]{%
1971 \let\hyxmp@textunderscore=\textunderscore
1972 \let\textunderscore=\hyxmp@uscore
1973 \pdfstringdef{#1}{#2}%
1974 \let\textunderscore=\hyxmp@textunderscore
1975 }
1976 % \end{macrocode}
1977 % \end{macro}
1978 % \end{macro}
1979 %
1980 % \begin{macro}{\@pdfdatetime}
1981 % Prepare to store the document's date and (optionally) time. Whether
1982 % specified by the author in \acro{XMP} format or \acro{PDF} format
1983 % (see Section~\ref{sec:date-manip}) we always store |\@pdfdatetime| as
1984 % an \acro{XMP}-format string.
1985 % \begin{macrocode}
1986 \def\@pdfdatetime{}
1987 \define@key{Hyp}{pdfdate}{%
1988 \begingroup
1989 \Hy@unicodefalse
1990 % \end{macrocode}
1991 % \begin{macro}{\next}
1992 % Expand \optname{pdfdate}'s argument and convert it to \acro{XMP} format.
1993 % \begin{macrocode}
1994 \edef\next{%
1995 \noexpand\hyxmp@pdfstringdef\noexpand\@pdfdatetime{%
1996 \noexpand\hyxmp@as@xmp@date{#1}}%
1997 }%
1998 \next
1999 \endgroup
2000 }
2001 % \end{macrocode}
2002 % \end{macro}
2003 % \end{macro}
2004 %
2005 % \begin{macro}{\@pdfmetadatetime}
2006 % Prepare to store the document's metadata date and (optionally) time.
2007 % Whether specified by the author in \acro{XMP} format or \acro{PDF}
2008 % format (see Section~\ref{sec:date-manip}) we always store
2009 % |\@pdfmetadatetime| as an \acro{XMP}-format string.
2010 % \begin{macrocode}
2011 \def\@pdfmetadatetime{}
2012 \define@key{Hyp}{pdfmetadate}{%
2013 \begingroup
2014 \Hy@unicodefalse
2015 % \end{macrocode}
2016 % \begin{macro}{\next}
2017 % Expand \optname{pdfmetadate}'s argument and convert it to \acro{XMP} format.
2018 % \begin{macrocode}
2019 \edef\next{%
2020 \noexpand\hyxmp@pdfstringdef\noexpand\@pdfmetadatetime{%
2021 \noexpand\hyxmp@as@xmp@date{#1}}%
2022 }%
2023 \next
2024 \endgroup
2025 }
2026 % \end{macrocode}
2027 % \end{macro}
2028 % \end{macro}
2029 %
2030 % \begin{macro}{\@pdfcopyright}
2031 % Prepare to store the document's copyright statement.
2032 % \begin{macrocode}
2033 \def\@pdfcopyright{}
2034 \define@key{Hyp}{pdfcopyright}{\hyxmp@pdfstringdef\@pdfcopyright{#1}}
2035 % \end{macrocode}
2036 % \end{macro}
2037 %
2038 % \begin{macro}{\@pdftype}
2039 % Prepare to store the document's logical type, which defaults to ``|Text|''.
2040 % \begin{macrocode}
2041 \def\@pdftype{Text}
2042 \define@key{Hyp}{pdftype}{\hyxmp@pdfstringdef\@pdftype{#1}}
2043 % \end{macrocode}
2044 % \end{macro}
2045 %
2046 % \begin{macro}{\@pdflicenseurl}
2047 % Prepare to store the \acro{URL} containing the document's license
2048 % agreement.
2049 % \begin{macrocode}
2050 \def\@pdflicenseurl{}
2051 \define@key{Hyp}{pdflicenseurl}{\hyxmp@pdfstringdef\@pdflicenseurl{#1}}
2052 % \end{macrocode}
2053 % \end{macro}
2054 %
2055 % \begin{macro}{\@pdfauthortitle}
2056 % Prepare to store the author's position/title (e.g.,~Staff Writer).
2057 % \begin{macrocode}
2058 \def\@pdfauthortitle{}
2059 \define@key{Hyp}{pdfauthortitle}{\hyxmp@pdfstringdef\@pdfauthortitle{#1}}
2060 % \end{macrocode}
2061 % \end{macro}
2062 %
2063 % \begin{macro}{\@pdfcaptionwriter}
2064 % Prepare to store the name of the person who inserted the
2065 % \pkgname{hyperxmp} metadata.
2066 % \begin{macrocode}
2067 \def\@pdfcaptionwriter{}
2068 \define@key{Hyp}{pdfcaptionwriter}{\hyxmp@pdfstringdef\@pdfcaptionwriter{#1}}
2069 % \end{macrocode}
2070 % \end{macro}
2071 %
2072 % \begin{macro}{\@pdfmetalang}
2073 % Prepare to store the natural language of the document's metadata,
2074 % typically as an \acro{ISO}~\mbox{639-1} two-letter abbreviation.
2075 % \begin{macrocode}
2076 \def\@pdfmetalang{}
2077 \define@key{Hyp}{pdfmetalang}{\hyxmp@pdfstringdef\@pdfmetalang{#1}}
2078 % \end{macrocode}
2079 % \end{macro}
2080 %
2081 % \begin{macro}{\hyxmp@no@bad@parts}
2082 % Complain about a bad \optname{pdfapart} or \optname{pdfuapart} if given
2083 % trailing non-digits after a part number.
2084 % \begin{macrocode}
2085 \def\hyxmp@no@bad@parts#1\relax{%
2086 \@ifnotmtarg{#1}{%
2087 \PackageWarning{hyperxmp}{pdfapart and pdfuapart must be numeric}%
2088 }%
2089 }
2090 % \end{macrocode}
2091 % \end{macro}
2092 %
2093 % \begin{macro}{\@hyxmp@count}
2094 % \changes{v5.11}{2022/10/02}{Added this macro to fix a bug with
2095 % \protect\optname{pdfapart}. Thanks to John H. Lienhard and
2096 % Kartik Singhal for their bug reports}
2097 % Define a temporary counter. The code previously used \cs{@tempcnta},
2098 % but this is no longer safe within \cs{pdfstringdef} as of more recent
2099 % versions of \pkgname{hyperref}.
2100 % \begin{macrocode}
2101 \newcount\@hyxmp@count
2102 % \end{macrocode}
2103 % \end{macro}
2104 %
2105 % \begin{macro}{\@pdfapart}
2106 % Prepare to store the \acro{PDF/A} part ID, which defaults to~``1''
2107 % if \optname{pdfa} is passed to \pkgname{hyperref}.
2108 % \begin{macrocode}
2109 \def\@pdfapart{}
2110 \define@key{Hyp}{pdfapart}{%
2111 \afterassignment\hyxmp@no@bad@parts\@hyxmp@count=0#1\relax
2112 \hyxmp@pdfstringdef\@pdfapart{\the\@hyxmp@count}%
2113 }
2114 % \end{macrocode}
2115 % \end{macro}
2116 %
2117 % \begin{macro}{\@pdfaconformance}
2118 % Prepare to store the \acro{PDF/A} conformance ID, which defaults to~``b''
2119 % if \optname{pdfa} is passed to \pkgname{hyperref} and |\@pdfapart| is
2120 % empty.
2121 % \begin{macrocode}
2122 \def\@pdfaconformance{}
2123 \define@key{Hyp}{pdfaconformance}{%
2124 \uppercase{\hyxmp@pdfstringdef\@pdfaconformance{#1}}%
2125 }
2126 % \end{macrocode}
2127 % \end{macro}
2128 %
2129 % \begin{macro}{\@pdfuapart}
2130 % Prepare to store the \acro{PDF/UA} part ID.
2131 % \begin{macrocode}
2132 \def\@pdfuapart{}
2133 \define@key{Hyp}{pdfuapart}{%
2134 \afterassignment\hyxmp@no@bad@parts\@hyxmp@count=0#1\relax
2135 \hyxmp@pdfstringdef\@pdfuapart{\the\@hyxmp@count}%
2136 }
2137 % \end{macrocode}
2138 % \end{macro}
2139 %
2140 % \begin{macro}{\hyxmp@set@pdfx@major}
2141 % Parse \optname{pdfxstandard} as ``|PDF/X-|\meta{major}\meta{other}'',
2142 % setting |\hyxmp@pdfx@major| to \meta{major}.
2143 % \begin{macrocode}
2144 \newcommand*{\hyxmp@set@pdfx@major}[1]{\hyxmp@set@pdfx@major@i#1!}
2145 % \end{macrocode}
2146 % \end{macro}
2147 %
2148 % \begin{macro}{\hyxmp@set@pdfx@major@i}
2149 % This is the first helper macro for |\hyxmp@set@pdfx@major|. It stores
2150 % the \acro{PDF/X} major version in |\@hyxmp@count|.
2151 % \begin{macrocode}
2152 \def\hyxmp@set@pdfx@major@i PDF/X-{%
2153 \afterassignment\hyxmp@set@pdfx@major@ii
2154 \@hyxmp@count=%
2155 }
2156 % \end{macrocode}
2157 % \end{macro}
2158 %
2159 % \begin{macro}{\hyxmp@set@pdfx@major@ii}
2160 % \begin{macro}{\hyxmp@pdfx@major}
2161 % This is the second helper macro for |\hyxmp@set@pdfx@major|. It
2162 % copies the \acro{PDF/X} major version from |\@hyxmp@count| to
2163 % |\@hyxmp@pdfx@major| and discards the rest of the \acro{PDF/X}
2164 % standard string.
2165 % \begin{macrocode}
2166 \def\hyxmp@set@pdfx@major@ii#1!{%
2167 \edef\hyxmp@pdfx@major{\the\@hyxmp@count}%
2168 }
2169 % \end{macrocode}
2170 % \end{macro}
2171 % \end{macro}
2172 %
2173 % \begin{macro}{\hyxmp@check@std}
2174 % \changes{v5.0}{2020/03/10}{Added this macro}
2175 % Compare a user-provided string to a fixed string. (Assumption: Both
2176 % are names of \acro{PDF/X} standard versions.) If they match, undefine
2177 % |\next|, which we assume was previously defined to issue an
2178 % ``unrecognized standard'' warning message.
2179 % \begin{macrocode}
2180 \newcommand*\hyxmp@check@std[2]{%
2181 \ifthenelse{\equal{#1}{#2}}%
2182 {\global\let\next=\relax}%
2183 {}%
2184 }%
2185 % \end{macrocode}
2186 % \end{macro}
2187 %
2188 % \begin{macro}{\@pdfxstandard}
2189 % Prepare to store the \acro{PDF/X} standard.
2190 % \changes{v5.0}{2020/03/10}{Added this macro}
2191 % \begin{macrocode}
2192 \def\@pdfxstandard{}
2193 \def\hyxmp@pdfx@major{}
2194 \define@key{Hyp}{pdfxstandard}{%
2195 \hyxmp@pdfstringdef\@pdfxstandard{#1}%
2196 % \end{macrocode}
2197 % \begin{macro}{\next}
2198 % Issue a warning message if the \acro{PDF/X} standard named by the user
2199 % does not appear in a list of known \acro{PDF/X} standards. This is to
2200 % caution the user that \pkgname{hyperxmp} generates standard-specific
2201 % \acro{XMP} metadata and it can only guess at the correct format for
2202 % new standard versions. (See the comments on
2203 % page~\pageref{page:pdfx-id-schema} above the definition of
2204 % |\hyxmp@pdfx@id@schema|, for example.)
2205 % \begin{macrocode}
2206 \gdef\next{%
2207 \PackageWarning{hyperxmp}{Unrecognized PDF/X standard `#1'}%
2208 }%
2209 \hyxmp@check@std{#1}{PDF/X-1a:2001}%
2210 \hyxmp@check@std{#1}{PDF/X-1a:2003}%
2211 \hyxmp@check@std{#1}{PDF/X-3:2002}%
2212 \hyxmp@check@std{#1}{PDF/X-3:2003}%
2213 \hyxmp@check@std{#1}{PDF/X-4}%
2214 \hyxmp@check@std{#1}{PDF/X-4p}%
2215 \hyxmp@check@std{#1}{PDF/X-5g}%
2216 \hyxmp@check@std{#1}{PDF/X-5n}%
2217 \hyxmp@check@std{#1}{PDF/X-5pg}%
2218 \next
2219 % \end{macrocode}
2220 % \begin{macro}{\hyxmp@pdfx@major}
2221 % Parse the \acro{PDF/X} major version number from
2222 % \optname{pdfxstandard} and assign it to |\hyxmp@pdfx@major|.
2223 % \begin{macrocode}
2224 \hyxmp@set@pdfx@major{#1}%
2225 }
2226 % \end{macrocode}
2227 % \end{macro}
2228 % \end{macro}
2229 % \end{macro}
2230 %
2231 % \begin{macro}{\@pdfsource}
2232 % Prepare to store the document's source, which defaults to the value of
2233 % |\jobname|.
2234 % \changes{v3.3}{2017/07/16}{Added this macro and the corresponding
2235 % \protect\optname{pdfsource} option, at Niklas Beisert's request}
2236 % \begin{macrocode}
2237 \edef\@pdfsource{\[email protected]}
2238 \define@key{Hyp}{pdfsource}{\hyxmp@pdfstringdef\@pdfsource{#1}}
2239 % \end{macrocode}
2240 % \end{macro}
2241 %
2242 % \begin{macro}{\hyxmp@DocumentID}
2243 % Prepare to store a \acro{UUID} that represents the document.
2244 % \changes{v3.5}{2018/11/27}{Added the \protect\optname{pdfdocumentid}
2245 % option, at Michael Osipov's request}
2246 % \begin{macrocode}
2247 \def\hyxmp@DocumentID{}
2248 \define@key{Hyp}{pdfdocumentid}{\hyxmp@pdfstringdef\hyxmp@DocumentID{#1}}
2249 % \end{macrocode}
2250 % \end{macro}
2251 %
2252 % \begin{macro}{\hyxmp@InstanceID}
2253 % Prepare to store a \acro{UUID} that represents the current instance of
2254 % the document.
2255 % \changes{v3.5}{2018/11/27}{Added the \protect\optname{pdfinstanceid}
2256 % option, at Michael Osipov's request}
2257 % \begin{macrocode}
2258 \def\hyxmp@InstanceID{}
2259 \define@key{Hyp}{pdfinstanceid}{\hyxmp@pdfstringdef\hyxmp@InstanceID{#1}}
2260 % \end{macrocode}
2261 % \end{macro}
2262 %
2263 % \begin{macro}{\@pdfversionid}
2264 % Prepare to store a string that represents the current version of
2265 % the document. It defaults to ``|1|''.
2266 % \begin{macrocode}
2267 \def\@pdfversionid{1}
2268 \define@key{Hyp}{pdfversionid}{\hyxmp@pdfstringdef\@pdfversionid{#1}}
2269 % \end{macrocode}
2270 % \end{macro}
2271 %
2272 % \begin{macro}{\ifdraft}
2273 % \changes{v5.2}{2020/04/18}{Define \protect\cs{ifdraft} only locally, at
2274 % Niklas Beisert's request}
2275 % \changes{v5.6}{2020/10/05}{Make conditional the loading of the
2276 % \protect\pkgname{ifdraft} package. Thanks to Tobias Pape for
2277 % reporting the incompatibility between \protect\pkgname{hyperxmp} and
2278 % \protect\pkgname{ifdraft}}
2279 % \changes{v5.13}{2024/03/17}{Rewrite \protect\pkgname{ifdraft} handling
2280 % to avoid loading a package within a group, which \protect\LaTeX\
2281 % soon will stop supporting. Thanks to Ulrike Fischer and Boris Veytsman
2282 % for bringing this to my attention}
2283 % \begin{macro}{\next}
2284 % Use the \pkgname{ifdraft} package to determine if this is a draft or
2285 % final document. The challenge here is that we want to use
2286 % \pkgname{ifdraft} if it's already loaded, load it if not, and not
2287 % break any incompatible, author-defined \cs{ifdraft} macros that may
2288 % occur either before or after the |\usepackage{hyperxmp}|. Our
2289 % solution works as follows:
2290 %
2291 % \begin{itemize}
2292 % \item If \pkgname{ifdraft} is already loaded, \cs{next} is defined
2293 % as a no-op.
2294 % \item If \pkgname{ifdraft} is not (yet) loaded, the code backs up
2295 % then undefines \cs{ifdraft}, which may be author-defined. It then
2296 % loads \pkgname{ifdraft} and defines \cs{next} to ``unload'' the
2297 % package by setting the package's author-visible commands to
2298 % \cs{relax}.
2299 % \item Below, after \cs{ifdraft} is used to define
2300 % \cs{@pdfrendition}, \cs{next} is invoked to unload a
2301 % \pkgname{hyperxmp}-loaded \pkgname{ifdraft}. Finally, \cs{ifdraft}
2302 % is restored to its original definition.
2303 % \end{itemize}
2304 % \begin{macrocode}
2305 \let\hyxmp@orig@ifdraft=\ifdraft
2306 \@ifpackageloaded{ifdraft}{%
2307 \let\next=\relax
2308 }{%
2309 \let\ifdraft=\relax
2310 \RequirePackage{ifdraft}%
2311 \def\next{%
2312 \expandafter\let\csname [email protected]\endcsname=\relax
2313 \let\ifdraft=\relax
2314 \let\ifoptiondraft=\relax
2315 \let\ifoptionfinal=\relax
2316 }%
2317 }%
2318 % \end{macrocode}
2319 % \begin{macro}{\@pdfrendition}
2320 % \changes{v5.0}{2020/02/27}{Added the \protect\optname{pdfrendition}
2321 % option}
2322 % Prepare to store a tag describing how this rendition of the document
2323 % differs from the master. The default value is |default|, which
2324 % indicates the master document, except in the case of
2325 % |\documentclass[draft]|, for which |\@pdfrendition| defaults to
2326 % |draft|.
2327 % \begin{macrocode}
2328 \ifdraft{%
2329 \def\@pdfrendition{draft}%
2330 }{%
2331 \def\@pdfrendition{default}%
2332 }
2333 \next
2334 \let\ifdraft=\hyxmp@orig@ifdraft
2335 \define@key{Hyp}{pdfrendition}{\hyxmp@pdfstringdef\@pdfrendition{#1}}
2336 % \end{macrocode}
2337 % \end{macro}
2338 % \end{macro}
2339 % \end{macro}
2340 %
2341 % \begin{macro}{\@pdfpublication}
2342 % Prepare to store the name of the publication in which the
2343 % document was published.
2344 % \begin{macrocode}
2345 \def\@pdfpublication{}
2346 \define@key{Hyp}{pdfpublication}{\hyxmp@pdfstringdef\@pdfpublication{#1}}
2347 % \end{macrocode}
2348 % \end{macro}
2349 %
2350 % \begin{macro}{\@pdfpubtype}
2351 % Prepare to store the type of the publication in which the
2352 % document was published.
2353 % \begin{macrocode}
2354 \def\@pdfpubtype{}
2355 \define@key{Hyp}{pdfpubtype}{\hyxmp@pdfstringdef\@pdfpubtype{#1}}
2356 % \end{macrocode}
2357 % \end{macro}
2358 %
2359 % \begin{macro}{\@pdfbytes}
2360 % Prepare to store the size of the file in bytes.
2361 % \begin{macrocode}
2362 \def\@pdfbytes{}
2363 \define@key{Hyp}{pdfbytes}{\hyxmp@pdfstringdef\@pdfbytes{#1}}
2364 % \end{macrocode}
2365 % \end{macro}
2366 %
2367 % \begin{macro}{\@pdfnumpages}
2368 % Prepare to store the number of pages in the file.
2369 % \begin{macrocode}
2370 \def\@pdfnumpages{}
2371 \define@key{Hyp}{pdfnumpages}{\hyxmp@pdfstringdef\@pdfnumpages{#1}}
2372 % \end{macrocode}
2373 % \end{macro}
2374 %
2375 % \begin{macro}{\@pdfissn}
2376 % Prepare to store the \acro{ISSN} of the publication in which the
2377 % document was published.
2378 % \begin{macrocode}
2379 \def\@pdfissn{}
2380 \define@key{Hyp}{pdfissn}{\hyxmp@pdfstringdef\@pdfissn{#1}}
2381 % \end{macrocode}
2382 % \end{macro}
2383 %
2384 % \begin{macro}{\@pdfeissn}
2385 % Prepare to store the \acro{ISSN} of the electronic version of the
2386 % publication in which the document was published.
2387 % \begin{macrocode}
2388 \def\@pdfeissn{}
2389 \define@key{Hyp}{pdfeissn}{\hyxmp@pdfstringdef\@pdfeissn{#1}}
2390 % \end{macrocode}
2391 % \end{macro}
2392 %
2393 % \begin{macro}{\@pdfisbn}
2394 % Prepare to store the \acro{ISBN} of the publication in which the
2395 % document was published.
2396 % \begin{macrocode}
2397 \def\@pdfisbn{}
2398 \define@key{Hyp}{pdfisbn}{\hyxmp@pdfstringdef\@pdfisbn{#1}}
2399 % \end{macrocode}
2400 % \end{macro}
2401 %
2402 % \begin{macro}{\@pdfbookedition}
2403 % Prepare to store the edition of the book in which the document
2404 % was published.
2405 % \begin{macrocode}
2406 \def\@pdfbookedition{}
2407 \define@key{Hyp}{pdfbookedition}{\hyxmp@pdfstringdef\@pdfbookedition{#1}}
2408 % \end{macrocode}
2409 % \end{macro}
2410 %
2411 % \begin{macro}{\@pdfpublisher}
2412 % Prepare to store the name of the document's publisher.
2413 % \begin{macrocode}
2414 \def\@pdfpublisher{}
2415 \define@key{Hyp}{pdfpublisher}{\hyxmp@pdfstringdef\@pdfpublisher{#1}}
2416 % \end{macrocode}
2417 % \end{macro}
2418 %
2419 % \begin{macro}{\@pdfvolumenum}
2420 % Prepare to store the volume identifier of the publication in which the
2421 % document was published.
2422 % \begin{macrocode}
2423 \def\@pdfvolumenum{}
2424 \define@key{Hyp}{pdfvolumenum}{\hyxmp@pdfstringdef\@pdfvolumenum{#1}}
2425 % \end{macrocode}
2426 % \end{macro}
2427 %
2428 % \begin{macro}{\@pdfissuenum}
2429 % Prepare to store the identifier of the issue within a volume of the
2430 % publication in which the document was published.
2431 % \begin{macrocode}
2432 \def\@pdfissuenum{}
2433 \define@key{Hyp}{pdfissuenum}{\hyxmp@pdfstringdef\@pdfissuenum{#1}}
2434 % \end{macrocode}
2435 % \end{macro}
2436 %
2437 % \begin{macro}{\@pdfpagerange}
2438 % Prepare to store the document's range of pages within the publication
2439 % in which the document was published.
2440 % \begin{macrocode}
2441 \def\@pdfpagerange{}
2442 \define@key{Hyp}{pdfpagerange}{\hyxmp@pdfstringdef\@pdfpagerange{#1}}
2443 % \end{macrocode}
2444 % \end{macro}
2445 %
2446 % \begin{macro}{\@pdfdoi}
2447 % Prepare to store a \acro{DOI} that represents the current instance of
2448 % the document.
2449 % \begin{macrocode}
2450 \def\@pdfdoi{}
2451 \define@key{Hyp}{pdfdoi}{\hyxmp@pdfstringdef\@pdfdoi{#1}}
2452 % \end{macrocode}
2453 % \end{macro}
2454 %
2455 % \begin{macro}{\@pdfurl}
2456 % Prepare to store a \acro{URL} that represents where the document
2457 % can be found. Note that we do not prepend \optname{baseurl} to
2458 % the value provided.
2459 % \begin{macrocode}
2460 \def\@pdfurl{}
2461 \define@key{Hyp}{pdfurl}{\hyxmp@pdfstringdef\@pdfurl{#1}}
2462 % \end{macrocode}
2463 % \end{macro}
2464 %
2465 % \begin{macro}{\@pdfidentifier}
2466 % Prepare to store an identifier that uniquely represents the document.
2467 % \begin{macrocode}
2468 \def\@pdfidentifier{}
2469 \define@key{Hyp}{pdfidentifier}{\hyxmp@pdfstringdef\@pdfidentifier{#1}}
2470 % \end{macrocode}
2471 % \end{macro}
2472 %
2473 % \begin{macro}{\@pdfsubtitle}
2474 % Prepare to store the document's subtitle.
2475 % \begin{macrocode}
2476 \def\@pdfsubtitle{}
2477 \define@key{Hyp}{pdfsubtitle}{\hyxmp@pdfstringdef\@pdfsubtitle{#1}}
2478 % \end{macrocode}
2479 % \end{macro}
2480 %
2481 % \begin{macro}{\@pdfpubstatus}
2482 % Prepare to store the document's journal article version.
2483 % \begin{macrocode}
2484 \def\@pdfpubstatus{}
2485 \define@key{Hyp}{pdfpubstatus}{\hyxmp@pdfstringdef\@pdfpubstatus{#1}}
2486 % \end{macrocode}
2487 % \end{macro}
2488 %
2489 % The following eight macros---|\@pdfcontactaddress|,
2490 % |\@pdfcontactcity|, |\@pdfcontactregion|, |\@pdfcontactpostcode|,
2491 % |\@pdfcontactcountry|, |\@pdfcontactphone|, |\@pdfcontactemail|, and
2492 % |\@pdfcontacturl|---together specify how to contact the person or
2493 % institution responsible for the document.
2494 %
2495 % \begin{macro}{\@pdfcontactaddress}
2496 % Prepare to store a street address for the document's contact
2497 % person\slash institution. The \acro{IPTC} standard defines this as
2498 % follows:
2499 %
2500 % \begin{quote}
2501 % The contact information address part. Comprises an optional company
2502 % name and all required information to locate the building or postbox
2503 % to which mail should be sent. To that end, the address is a
2504 % multiline field.
2505 % \end{quote}
2506 %
2507 % For consistency with the rest of \pkgname{hyperxmp}, we use commas to
2508 % separate terms, in this case, lines of the address. The author can
2509 % use |\xmpquote| and |\xmpcomma| to include literal commas.
2510 % \begin{macrocode}
2511 \def\@pdfcontactaddress{}
2512 \define@key{Hyp}{pdfcontactaddress}{%
2513 \let\xmpcomma=\hyxmp@comma
2514 \def\xmpquote##1{##1}%
2515 \hyxmp@pdfstringdef\@pdfcontactaddress{#1}%
2516 \def\xmpcomma{,}%
2517 \let\xmpquote=\relax
2518 }
2519 % \end{macrocode}
2520 % \end{macro}
2521 %
2522 % \begin{macro}{\@pdfcontactcity}
2523 % Prepare to store the city of the document's contact person\slash
2524 % institution.
2525 % \begin{macrocode}
2526 \def\@pdfcontactcity{}
2527 \define@key{Hyp}{pdfcontactcity}{\hyxmp@pdfstringdef\@pdfcontactcity{#1}}
2528 % \end{macrocode}
2529 % \end{macro}
2530 %
2531 % \begin{macro}{\@pdfcontactregion}
2532 % Prepare to store the state or province of the document's contact
2533 % person\slash institution.
2534 % \begin{macrocode}
2535 \def\@pdfcontactregion{}
2536 \define@key{Hyp}{pdfcontactregion}{\hyxmp@pdfstringdef\@pdfcontactregion{#1}}
2537 % \end{macrocode}
2538 % \end{macro}
2539 %
2540 % \begin{macro}{\@pdfcontactpostcode}
2541 % Prepare to store the postal code of the document's contact
2542 % person\slash institution.
2543 % \begin{macrocode}
2544 \def\@pdfcontactpostcode{}
2545 \define@key{Hyp}{pdfcontactpostcode}{\hyxmp@pdfstringdef\@pdfcontactpostcode{#1}}
2546 % \end{macrocode}
2547 % \end{macro}
2548 %
2549 % \begin{macro}{\@pdfcontactcountry}
2550 % Prepare to store the country of the document's contact person\slash
2551 % institution.
2552 % \begin{macrocode}
2553 \def\@pdfcontactcountry{}
2554 \define@key{Hyp}{pdfcontactcountry}{\hyxmp@pdfstringdef\@pdfcontactcountry{#1}}
2555 % \end{macrocode}
2556 % \end{macro}
2557 %
2558 % \begin{macro}{\@pdfcontactphone}
2559 % Prepare to store the telephone number of the document's contact
2560 % person\slash institution.
2561 % \begin{macrocode}
2562 \def\@pdfcontactphone{}
2563 \define@key{Hyp}{pdfcontactphone}{\hyxmp@pdfstringdef\@pdfcontactphone{#1}}
2564 % \end{macrocode}
2565 % \end{macro}
2566 %
2567 % \begin{macro}{\@pdfcontactemail}
2568 % Prepare to store the email address of the document's contact
2569 % person\slash institution.
2570 % \begin{macrocode}
2571 \def\@pdfcontactemail{}
2572 \define@key{Hyp}{pdfcontactemail}{\hyxmp@pdfstringdef\@pdfcontactemail{#1}}
2573 % \end{macrocode}
2574 % \end{macro}
2575 %
2576 % \begin{macro}{\@pdfcontacturl}
2577 % Prepare to store the \acro{URL} of the document's contact person\slash
2578 % institution.
2579 % \begin{macrocode}
2580 \def\@pdfcontacturl{}
2581 \define@key{Hyp}{pdfcontacturl}{\hyxmp@pdfstringdef\@pdfcontacturl{#1}}
2582 % \end{macrocode}
2583 % \end{macro}
2584 %
2585 % \begin{macro}{\hyxmp@no@info@lists}
2586 % Suppress \pkgname{hyperref} from writing \pdfterm{Author} and
2587 % \pdfterm{Keywords} into the \pdfterm{Info} dictionary. This prevents
2588 % conflicts between the \acro{PDF} metadata and the \acro{XMP} metadata
2589 % that cause \acro{PDF/A} validation to fail. The \acro{PDF} metadata
2590 % can be restored by passing the \optname{keeppdfinfo} option to
2591 % |\hypersetup|.
2592 % \changes{v4.0}{2019/03/31}{Added this macro}
2593 % \changes{v5.0}{2020/03/10}{Renamed this macros from
2594 % \protect\cs{hyxmp@suppress@pdf@metadata} and rewrote it to replace,
2595 % if possible, only \protect\pdfterm{Author} and \protect\pdfterm{Keywords}}
2596 % \begin{macrocode}
2597 \def\hyxmp@no@info@lists{%
2598 % \end{macrocode}
2599 % \begin{macro}{\hyxmp@suppress@pdf@info}
2600 % \begin{macro}{\next}
2601 % If |\patchcmd| fails for any reason---most likely, a modification to
2602 % the \pkgname{hyperref} package---our fallback is to prevent
2603 % \pkgname{hyperref} from writing \emph{any} data to the \acro{PDF}
2604 % \pdfterm{Info} dictionary.
2605 % \begin{macrocode}
2606 \def\hyxmp@suppress@pdf@info{%
2607 \global\let\PDF@FinishDoc=\@empty
2608 \PackageWarningNoLine{hyperxmp}{%
2609 Suppressing the _entire_ PDF Info dictionary.\MessageBreak
2610 Please notify the hyperxmp maintainer%
2611 }%
2612 }%
2613 \let\next=\relax
2614 \patchcmd
2615 {\PDF@FinishDoc}%
2616 {/Author(\@pdfauthor)}%
2617 {}%
2618 {}%
2619 {\let\next=\hyxmp@suppress@pdf@info}%
2620 \patchcmd
2621 {\PDF@FinishDoc}%
2622 {/Keywords(\@pdfkeywords)}%
2623 {}%
2624 {}%
2625 {\let\next=\hyxmp@suppress@pdf@info}%
2626 \next
2627 }
2628 % \end{macrocode}
2629 % \end{macro}
2630 % \end{macro}
2631 % \end{macro}
2632 % \begin{macrocode}
2633 \define@key{Hyp}{keeppdfinfo}[true]{%
2634 \gdef\hyxmp@no@info@lists{}%
2635 }
2636 % \end{macrocode}
2637 %
2638 % \changes{v2.1}{2012/09/16}{Enabled \protect\pkgname{hyperxmp} and
2639 % \protect\pkgname{hyperref} to be loaded in either order. This addresses
2640 % a bug report by Yury Donskoy}
2641 %
2642 % We need to capture list arguments (viz.~\optname{pdfauthor} and
2643 % \optname{pdfkeywords}) before \pkgname{hyperref} converts them to
2644 % \term{PDFDocEncoding}. Otherwise, |\xmpcomma| is permanently replaced
2645 % with a comma, and we lose our ability to change it to a
2646 % |\hyxmp@comma|. We therefore need to augment \pkgname{hyperref}'s
2647 % option processing with our own. Because \pkgname{hyperref} has not
2648 % yet been loaded we need to ensure that our augmentation gets loaded in
2649 % the future: after the |\usepackage{hyperref}| but before options are
2650 % passed to that package.
2651 %
2652 % For lack of a better approach, \pkgname{hyperxmp} redefines
2653 % |\ProcessKeyvalOptions| to alter the way \pkgname{hyperref} processes
2654 % \optname{pdfauthor} and \optname{pdfkeywords}. This is somewhat
2655 % heavy-handed as it gets executed for \emph{every} subsequently loaded
2656 % package that uses |\ProcessKeyvalOptions|, but at least it does what
2657 % we need. \pkgname{hyperxmp} also redefines |\hypersetup| to do the
2658 % same thing. This is required in case \pkgname{hyperref} is loaded
2659 % before \pkgname{hyperxmp}.
2660 %
2661 % \changes{v5.12}{2023/09/10}{Require that \pkgname{hyperref} be loaded
2662 % before \pkgname{hyperxmp}}
2663 % \begin{center}
2664 % \fbox{%
2665 % \begin{minipage}{0.75\linewidth}
2666 % \paragraph{New in v5.12}
2667 % \pkgname{hyperref} must be loaded \emph{before} \pkgname{hyperxmp}.
2668 % This is because recent changes in \pkgname{hyperref} and the
2669 % \LaTeX\ kernel prevent \pkgname{hyperxmp} from hooking into
2670 % \pkgname{hyperref}'s internals if \pkgname{hyperref} is loaded
2671 % first.
2672 % \end{minipage}%
2673 % }
2674 % \end{center}
2675 %
2676 % \begin{macrocode}
2677 \@ifpackageloaded{hyperref}{%
2678 }{%
2679 \PackageError{hyperxmp}%
2680 {hyperref must be loaded before hyperxmp}%
2681 {Recent versions of hyperref and the LaTeX kernel inhibit\MessageBreak
2682 hyperxmp's ability to hook into hyperref's internals unless\MessageBreak
2683 hyperref is loaded first.}
2684 }
2685 % \end{macrocode}
2686 %
2687 % \begin{macro}{\hyxmp@pdfauthor}
2688 % \begin{macro}{\hyxmp@pdfkeywords}
2689 % Prepare to store the name of the author and a list of keywords.
2690 % \begin{macrocode}
2691 \def\hyxmp@pdfauthor{}
2692 \def\hyxmp@pdfkeywords{}
2693 % \end{macrocode}
2694 % \end{macro}
2695 % \end{macro}
2696 %
2697 % \begin{macro}{\hyxmp@redefine@Hyp}
2698 % If not already redefined, redefine \pkgname{hyperref}'s
2699 % \optname{pdfauthor} and \optname{pdfkeywords} options to properly
2700 % handle |\xmpcomma| and |\xmpquote|.
2701 % \changes{v2.1}{2012/09/16}{Added this macro}
2702 % \begin{macrocode}
2703 \newcommand*{\hyxmp@redefine@Hyp}{%
2704 % \end{macrocode}
2705 % \begin{macro}{\hyxmp@Hyp@pdfauthor}
2706 % Store the old definition of |\KV@Hyp@pdfauthor| in
2707 % |\hyxmp@Hyp@pdfauthor|, but only if we see that |\KV@Hyp@pdfauthor| is
2708 % defined and |\hyxmp@Hyp@pdfauthor| isn't. Otherwise, we'd be defining
2709 % |\hyxmp@Hyp@pdfauthor| in terms of itself and creating an infinite
2710 % loop.
2711 % \begin{macrocode}
2712 \@ifundefined{KV@Hyp@pdfauthor}{}{%
2713 \@ifundefined{hyxmp@Hyp@pdfauthor}{%
2714 \expandafter\let\expandafter\hyxmp@Hyp@pdfauthor
2715 \csname KV@Hyp@pdfauthor\endcsname
2716 }{}%
2717 }%
2718 % \end{macrocode}
2719 % \Needspace{6\baselineskip}
2720 % \begin{macro}{\KV@Hyp@pdfauthor}
2721 % \begin{macro}{\xmpcomma}
2722 % \begin{macro}{\xmpquote}
2723 % \begin{macro}{\hyxmp@and}
2724 % \begin{macro}{\and}
2725 % \begin{macro}{\hyxmp@pdfauthor}
2726 % \begin{macro}{\@pdfauthor}
2727 % Redefine |\KV@Hyp@pdfauthor| to process its argument twice. The first
2728 % time, |\xmpcomma| is defined as a placeholder character
2729 % (|\hyxmp@comma|) and |\xmpquote| as the identity function. The result
2730 % is stored in |\hyxmp@pdfauthor| for use in structured lists (those
2731 % surrounding each entry with |<rdf:li>|). The second time, |\xmpcomma|
2732 % is defined as an ordinary comma, and |\xmpquote| is defined as a macro
2733 % that puts its argument within double quotes. The result is stored in
2734 % |\@pdfauthor| for use in unstructured lists (those in which the entire
2735 % list appears within a single pair of tags). In case
2736 % \optname{pdfauthor} is left unspecified and we copy |\author|'s
2737 % argument to \optname{pdfauthor}, we temporarily redefine |\and| as the
2738 % list separator when producing a structured list and as ``|and|'' when
2739 % producing an unstructured list.
2740 % \begin{macrocode}
2741 \define@key{Hyp}{pdfauthor}{%
2742 \let\xmpcomma=\hyxmp@comma
2743 \def\xmpquote####1{####1}%
2744 \let\hyxmp@and=\and
2745 \def\and{,}%
2746 \hyxmp@Hyp@pdfauthor{##1}%
2747 \global\let\hyxmp@pdfauthor=\@pdfauthor
2748 \def\and{and\space}%
2749 \def\xmpcomma{,}%
2750 \def\xmpquote####1{"####1"}%
2751 \hyxmp@Hyp@pdfauthor{##1}%
2752 \def\xmpcomma{,}%
2753 \let\xmpquote=\relax
2754 \let\and=\hyxmp@and
2755 }%
2756 % \end{macrocode}
2757 % \begin{macro}{\hyxmp@Hyp@pdfkeywords}
2758 % The previous block of code now repeats for the keyword list, starting
2759 % by storing the old definition of |\KV@Hyp@pdfkeywords| in
2760 % |\hyxmp@Hyp@pdfkeywords|.
2761 % \begin{macrocode}
2762 \@ifundefined{KV@Hyp@pdfkeywords}{}{%
2763 \@ifundefined{hyxmp@Hyp@pdfkeywords}{%
2764 \expandafter\let\expandafter\hyxmp@Hyp@pdfkeywords
2765 \csname KV@Hyp@pdfkeywords\endcsname
2766 }{}%
2767 }%
2768 % \end{macrocode}
2769 % \begin{macro}{\KV@Hyp@pdfkeywords}
2770 % \begin{macro}{\xmpcomma}
2771 % \begin{macro}{\xmpquote}
2772 % \begin{macro}{\hyxmp@pdfkeywords}
2773 % \begin{macro}{\@pdfkeywords}
2774 % Redefine |\KV@Hyp@pdfkeywords| to process its argument twice. The
2775 % first time, |\xmpcomma| is defined as a placeholder character
2776 % (|\hyxmp@comma|) and |\xmpquote| as the identity function. The
2777 % result is stored in |\hyxmp@pdfkeywords| for use in structured lists
2778 % (those surrounding each entry with |<rdf:li>|). The second time,
2779 % |\xmpcomma| is defined as an ordinary comma, and |\xmpquote| is
2780 % defined as a macro that puts its argument within double quotes. The
2781 % result is stored in |\@pdfkeywords| for use in unstructured lists
2782 % (those in which the entire list appears within a single pair of tags).
2783 % \begin{macrocode}
2784 \define@key{Hyp}{pdfkeywords}{%
2785 \let\xmpcomma=\hyxmp@comma
2786 \def\xmpquote####1{####1}%
2787 \hyxmp@Hyp@pdfkeywords{##1}%
2788 \global\let\hyxmp@pdfkeywords=\@pdfkeywords
2789 \def\xmpcomma{,}%
2790 \def\xmpquote####1{"####1"}%
2791 \hyxmp@Hyp@pdfkeywords{##1}%
2792 \def\xmpcomma{,}%
2793 \let\xmpquote=\relax
2794 }%
2795 }
2796 % \end{macrocode}
2797 % \end{macro}
2798 % \end{macro}
2799 % \end{macro}
2800 % \end{macro}
2801 % \end{macro}
2802 % \end{macro}
2803 % \end{macro}
2804 % \end{macro}
2805 % \end{macro}
2806 % \end{macro}
2807 % \end{macro}
2808 % \end{macro}
2809 % \end{macro}
2810 % \end{macro}
2811 % \end{macro}
2812 %
2813 % \begin{macro}{\hyxmp@ProcessKeyvalOptions}
2814 % \changes{v2.0}{2012/09/05}{Added this macro}
2815 % \changes{v5.11}{2022/10/02}{Bug fix: Restore
2816 % \protect\cs{ProcessKeyvalOptions} after first use. Thanks to
2817 % Ulrike Fischer for the bug report}
2818 % \begin{macro}{\ProcessKeyvalOptions}
2819 % \changes{v2.0}{2012/09/05}{Added this macro}
2820 % Redefine \pkgname{kvoptions}'s |\ProcessOptions| command to invoke
2821 % |\hyxmp@redefine@Hyp| before performing its normal option processing.
2822 % \begin{macrocode}
2823 \let\hyxmp@ProcessKeyvalOptions=\ProcessKeyvalOptions
2824 \renewcommand*{\ProcessKeyvalOptions}{%
2825 \global\let\ProcessKeyvalOptions=\hyxmp@ProcessKeyvalOptions
2826 \hyxmp@redefine@Hyp
2827 \hyxmp@ProcessKeyvalOptions
2828 }
2829 % \end{macrocode}
2830 % \end{macro}
2831 % \end{macro}
2832 %
2833 % \begin{macro}{\hyxmp@hypersetup}
2834 % \changes{v2.1}{2012/09/16}{Added this macro}
2835 % \begin{macro}{\hypersetup}
2836 % \changes{v2.1}{2012/09/16}{Added this macro}
2837 % Redefine \pkgname{hyperref}'s |\hypersetup| command to invoke
2838 % |\hyxmp@redefine@Hyp| before performing its normal option processing.
2839 % \begin{macrocode}
2840 \let\hyxmp@hypersetup=\hypersetup
2841 \def\hypersetup{%
2842 \hyxmp@redefine@Hyp
2843 \hyxmp@hypersetup
2844 }
2845 % \end{macrocode}
2846 % \end{macro}
2847 % \end{macro}
2848 %
2849 % \begin{macro}{\hyxmp@concated@metadata}
2850 % \begin{macro}{\hyxmp@aep@toks}
2851 % Assume that if the document loaded either \pkgname{babel} or
2852 % \pkgname{polyglossia} it will eventually define one or more languages
2853 % that \pkgname{hyperxmp} can list within a \xmpprop{dc:language}
2854 % element. As explained in Section~\ref{sec:initial-prep}, we defer the
2855 % invocation of \cs{AtEndPreamble} to the end of the file.
2856 % \begin{macrocode}
2857 \edef\hyxmp@concated@metadata{}
2858 \expandafter\hyxmp@aep@toks\expandafter=\expandafter{%
2859 \the\hyxmp@aep@toks
2860 \AtEndPreamble{%
2861 \@ifpackageloaded{babel}{%
2862 \edef\hyxmp@concated@metadata{babel}%
2863 }{%
2864 \@ifpackageloaded{polyglossia}{%
2865 \edef\hyxmp@concated@metadata{polyglossia}%
2866 }{%
2867 }%
2868 }%
2869 }%
2870 }
2871 % \end{macrocode}
2872 % \end{macro}
2873 % \end{macro}
2874 %
2875 % \begin{macro}{\hyxmp@warn@if@no@metadata}
2876 % \begin{macro}{\hyxmp@concated@metadata}
2877 % Issue a warning message if the author failed to specify any metadata
2878 % at all. This excludes metadata that is included automatically such as
2879 % the current timestamp. Note that we don't consider |\@pdfmetalang| as
2880 % metadata as that value is meaningful only when used in conjunction
2881 % with other information. We also don't examine |\@pdfapart| or
2882 % |\@pdfaconformance| because those have nonempty default values.
2883 % \begin{macrocode}
2884 \newcommand*{\hyxmp@warn@if@no@metadata}{%
2885 \edef\hyxmp@concated@metadata{%
2886 \hyxmp@concated@metadata
2887 \@baseurl
2888 \@pdfauthor
2889 \@pdfauthortitle
2890 \@pdfbookedition
2891 \@pdfbytes
2892 \@pdfcaptionwriter
2893 \@pdfcontactaddress
2894 \@pdfcontactcity
2895 \@pdfcontactcountry
2896 \@pdfcontactemail
2897 \@pdfcontactphone
2898 \@pdfcontactpostcode
2899 \@pdfcontactregion
2900 \@pdfcontacturl
2901 \@pdfcopyright
2902 \@pdfcreationdate
2903 \@pdfdatetime
2904 \@pdfdoi
2905 \@pdfeissn
2906 \@pdfidentifier
2907 \@pdfisbn
2908 \@pdfissn
2909 \@pdfissuenum
2910 \@pdfkeywords
2911 \@pdflang
2912 \@pdflicenseurl
2913 \@pdfmetadatetime
2914 \@pdfmoddate
2915 \@pdfnumpages
2916 \@pdfpagerange
2917 \@pdfpublication
2918 \@pdfpubtype
2919 \@pdfsubject
2920 \@pdfsubtitle
2921 \@pdftitle
2922 \@pdfuapart
2923 \@pdfurl
2924 \@pdfvolumenum
2925 \@pdfxstandard
2926 }%
2927 \ifx\hyxmp@concated@metadata\@empty
2928 \PackageWarningNoLine{hyperxmp}{%
2929 \[email protected] did not specify any metadata to\MessageBreak
2930 include in the XMP packet.\space\space Please see the\MessageBreak
2931 hyperxmp documentation for instructions on how to\MessageBreak
2932 provide metadata values to hyperxmp}%
2933 \fi
2934 }
2935 % \end{macrocode}
2936 % \end{macro}
2937 % \end{macro}
2938 %
2939 % \begin{macro}{\hyxmp@check@standards}
2940 % Most \acro{PDF} standards require that certain metadata be present.
2941 % If compliance with a \acro{PDF} standard is claimed but any of the
2942 % metadata it requires are absent, issue a warning message.
2943 % \changes{v5.0}{2020/03/11}{Added this macro}
2944 % \begin{macrocode}
2945 \newcommand*{\hyxmp@check@standards}{%
2946 % \end{macrocode}
2947 % If the \optname{pdfa} option was passed to \pkgname{hyperref} but
2948 % |\@pdfapart| is not set, set it to~|1| and |\@pdfaconformance| to~|B|.
2949 % \begin{macrocode}
2950 \ifHy@pdfa
2951 \@ifmtargexp{\@pdfapart}{%
2952 \PackageWarningNoLine{hyperxmp}{%
2953 `pdfa' was passed to hyperref, but `pdfapart' was\MessageBreak
2954 not specified.\space\space Setting pdfapart to `1' and\MessageBreak
2955 pdfaconformance to `B'%
2956 }%
2957 \gdef\@pdfapart{1}%
2958 \gdef\@pdfaconformance{B}%
2959 }%
2960 {}%
2961 \fi
2962 % \end{macrocode}
2963 %
2964 % \begin{macro}{\hyxmp@standards}
2965 % We define |\hyxmp@standards| to be non-empty if \emph{any} \acro{PDF}
2966 % standard is claimed (currently, \acro{PDF/A}, \acro{PDF/X}, or
2967 % \acro{PDF/UA}.
2968 % \begin{macrocode}
2969 \edef\hyxmp@standards{%
2970 \@pdfapart
2971 \@pdfxstandard
2972 \@pdfuapart
2973 }%
2974 % \end{macrocode}
2975 % Check that a document title was provided and is non-empty.
2976 % \begin{macrocode}
2977 \@ifnotmtargexp{\hyxmp@standards}{%
2978 \@ifmtargexp{\@pdftitle}{%
2979 \PackageWarningNoLine{hyperxmp}{%
2980 Missing pdftitle (required for PDF standards\MessageBreak
2981 compliance)%
2982 }%
2983 }%
2984 {}%
2985 }%
2986 }
2987 % \end{macrocode}
2988 % \end{macro}
2989 % \end{macro}
2990 %
2991 % \begin{macro}{\hyxmp@aep@toks}
2992 % Right before we reach the |\begin{document}| we check if
2993 % \pkgname{hyperref} was loaded. In normal usage, the document will
2994 % already have done a |\usepackage{hyperref}| because otherwise,
2995 % \cs{hypersetup} will not have been defined, and only a limited amount
2996 % of metadata will be included. However, in case the author is relying
2997 % exclusively on \pkgname{hyperxmp}'s automatically detected metadata,
2998 % we'll need to load \pkgname{hyperref} now. As explained in
2999 % Section~\ref{sec:initial-prep}, we defer the invocation of
3000 % \cs{AtEndPreamble} to the end of the file.
3001 % \changes{v5.5}{2020/06/24}{Load \protect\pkgname{hyperref} automatically
3002 % if the document does not do so explicitly, as requested by Robin Schwab}
3003 % \begin{macrocode}
3004 \expandafter\hyxmp@aep@toks\expandafter=\expandafter{%
3005 \the\hyxmp@aep@toks
3006 \AtEndPreamble{%
3007 \RequirePackage{hyperref}%
3008 % \end{macrocode}
3009 % Older versions of \pkgname{hyperref} write the \pdfterm{Info}
3010 % dictionary to the \acro{PDF} file at the end of the document. Newer
3011 % versions of \pkgname{hyperref} write the \pdfterm{Info} dictionary to
3012 % the \acro{PDF} file at the \emph{beginning} of the document. For
3013 % compatibility with both old and new \pkgname{hyperref} implementations
3014 % we suppress writing the \pdfterm{Info} dictionary here, at the
3015 % beginning of the document.
3016 % \changes{v4.1}{2019/04/02}{Invoke
3017 % \protect\cs{hyxmp@no@info@lists} at the beginning of the
3018 % document, for compatibility with both newer and older versions of
3019 % \protect\pkgname{hyperref}}
3020 % \begin{macrocode}
3021 \hyxmp@no@info@lists
3022 % \end{macrocode}
3023 % If \optname{pdftitle} is undefined but the author invoked \cs{title},
3024 % we copy the latter to the former. This addresses two problems:
3025 % (1)~handling \LaTeX\ classes in which \cs{maketitle} clears \cs{title}
3026 % and (2)~ensuring that \pkgname{hyperref} writes the same title to the
3027 % \acro{PDF} \pdfterm{Info} dictionary that \pkgname{hyperxmp} writes to
3028 % the \acro{XMP} packet. We do likewise for \cs{author}~$\rightarrow$
3029 % \optname{pdfauthor}.
3030 %
3031 % One tricky bit is that the standard \LaTeX\ classes do not define
3032 % \cs{@title} and \cs{@author} as empty strings but rather as calls to
3033 % \cs{@latex@warning@no@line} that complain about a missing
3034 % title/author. Hence, we can't simply test if the title and author are
3035 % empty because they're not. Instead, we first locally redefine
3036 % \cs{@latex@warning@no@line} to discard its argument then test if any
3037 % text remains.
3038 % \changes{v5.5}{2020/07/24}{Copy \protect\cs{title} to
3039 % \protect\optname{pdftitle} and \protect\cs{author} to
3040 % \protect\optname{pdfauthor} at the start of the document to improve
3041 % consistency between \protect\acro{XMP} and \protect\acro{PDF} metadata}
3042 % \begin{macrocode}
3043 \begingroup
3044 \let\@latex@warning@no@line=\@gobble
3045 \hyxmp@use@first@valid{pdftitle}{\@pdftitle}{%
3046 \scr@subject@var,%
3047 \@title
3048 }%
3049 \hyxmp@use@first@valid{pdfauthor}{\@pdfauthor}{%
3050 \scr@fromname@var,%
3051 \@author
3052 }%
3053 \endgroup
3054 }%
3055 }
3056 % \end{macrocode}
3057 % \end{macro}
3058 %
3059 % When we reach the |\end{document}| we need to gather up the metadata
3060 % specified explicitly by the user, infer additional metadata where
3061 % possible, and write the \acro{XMP} packet to the \acro{PDF} file.
3062 % \changes{v1.3}{2011/04/25}{Introduced the \protect\optname{pdfmetalang}
3063 % package option, which enables an author to specify the language in which he
3064 % wrote the document's metadata}
3065 % \changes{v2.0}{2012/08/02}{New \cs{AtBeginDocument} code from Heiko
3066 % Oberdiek to properly encode \cs{@pdfmetalang}}
3067 % \changes{v3.3}{2017/07/21}{Don't overwrite an existing
3068 % \protect\optname{pdfmetalang} with \protect\optname{pdflang} or
3069 % \protect\xmpterm{x-default}. This addresses a bug report by Niklas
3070 % Beisert}
3071 % \changes{v5.4}{2020/06/18}{Moved the automatic assignment of
3072 % \protect\cs{@pdflang} and \protect\cs{@pdfmetalang} from
3073 % \protect\cs{hyxmp@auto@assign@data} to within a call to
3074 % \protect\cs{hyxmp@at@end}}
3075 % \changes{v5.5}{2020/07/21}{Move most of the \protect\cs{AtEndPreamble}
3076 % code to \protect\cs{hyxmp@at@end}}
3077 % \begin{macrocode}
3078 \hyxmp@at@end{%
3079 % \end{macrocode}
3080 % Fill in any missing metadata we can using values provided by the author
3081 % via mechanisms other than the \cs{hypersetup} command.
3082 % \begin{macrocode}
3083 \hyxmp@auto@assign@data
3084 % \end{macrocode}
3085 % If the document claims to comply with one or more \acro{PDF}
3086 % standards, check that all of the requisite metadata are present.
3087 % \begin{macrocode}
3088 \hyxmp@check@standards
3089 % \end{macrocode}
3090 % We can finally construct the \acro{XMP} packet and write it to the
3091 % \acro{PDF} document catalog.
3092 % \begin{macrocode}
3093 \hyxmp@warn@if@no@metadata
3094 \hyxmp@embed@packet
3095 }
3096 % \end{macrocode}
3097 %
3098 %
3099 % \subsection{Advanced metadata detection}
3100 %
3101 % \pkgname{hyperxmp} strives to be as convenient and user-friendly as
3102 % possible. To that end, we try to automatically detect as much
3103 % metadata as possible. The author can of course augment or override
3104 % autodetected metadata by explicitly providing values to
3105 % \cs{hypersetup}, but the hope is that we can save the author some
3106 % effort in many cases.
3107 %
3108 % In this section, we identify additional metadata we can use. Most of
3109 % the functionality is class- or package-specific. For example, we
3110 % check for phone numbers provided to the \Koma\ letter classes via
3111 % |\setkomavar{fromphone}{|\dots|}| and/or
3112 % |\setkomavar{frommobilephone}{|\dots|}|, street addresses provided to
3113 % the \ACMclass\ article class via \cs{affiliation}, and languages
3114 % the \pkgname{polyglossia} package is instructed to load via
3115 % \cs{setdefaultlanguage} and \cs{setotherlanguage}.
3116 %
3117 % \begin{macro}{\hyxmp@set@koma@phones}
3118 % \changes{v5.3}{2020/06/07}{Added this macro}
3119 % \changes{v5.5}{2020/07/08}{Support hyperlinks and other markup in
3120 % \protect\texttt{frommobilephone} and \protect\texttt{fromphone},
3121 % as requested by Robin Schwab}
3122 % \begin{macro}{\hyxmp@koma@phones}
3123 % Define \cs{hyxmp@koma@phones} as a comma-separated list of the phone
3124 % numbers provided to a \Koma\ letter class (mobile and landline).
3125 % \begin{macrocode}
3126 \newcommand*{\hyxmp@set@koma@phones}{%
3127 \begingroup
3128 \Hy@unicodefalse
3129 \@if@def@and@nonempty{scr@frommobilephone@var}{%
3130 \@if@def@and@nonempty{scr@fromphone@var}{%
3131 \hyxmp@pdfstringdef\hyxmp@koma@phones{\scr@frommobilephone@var,\scr@fromphone@var}%
3132 }{%
3133 \hyxmp@pdfstringdef\hyxmp@koma@phones{\scr@frommobilephone@var}%
3134 }%
3135 }{%
3136 \@if@def@and@nonempty{scr@fromphone@var}{%
3137 \hyxmp@pdfstringdef\hyxmp@koma@phones{\scr@fromphone@var}%
3138 }{%
3139 }%
3140 }%
3141 \endgroup
3142 }
3143 % \end{macrocode}
3144 % \end{macro}
3145 % \end{macro}
3146 %
3147 % \begin{macro}{\hyxmp@use@first@valid}
3148 % Given a \pkgname{hyperxmp} option~(|#1|), its current value~(|#2|),
3149 % and a comma-separated list of option names~(|#3|), if the current
3150 % value is empy, invoke \cs{hypersetup} to set the option to the first
3151 % non-empty item in the list. If all items in the list are empty, do
3152 % nothing.
3153 % \changes{v5.3}{2020/05/29}{Added this macro}
3154 % \begin{macrocode}
3155 \newcommand*{\hyxmp@use@first@valid}[3]{%
3156 \@ifmtargexp{#2}{%
3157 \hyxmp@use@first@valid@i{#1}#3,!,%
3158 }%
3159 {}%
3160 }
3161 % \end{macrocode}
3162 % \end{macro}
3163 %
3164 % \begin{macro}{\hyxmp@use@first@valid@i}
3165 % This macro performs all the work for \cs{hyxmp@use@first@valid}. It
3166 % loops over a comma-separated list of macros~(|#2|), stopping when it
3167 % encounters an end-of-list marker~(``|!|''). The first list element
3168 % that is neither undefined nor empty is assigned to a given option
3169 % name~(|#1|) using \cs{hypersetup}.
3170 % \begin{macrocode}
3171 \def\hyxmp@use@first@valid@i#1#2,{%
3172 \def\next{\hyxmp@use@first@valid@i{#1}}%
3173 \ifx#2!%
3174 \let\next=\relax
3175 \else
3176 \ifx#2\undefined
3177 \else
3178 \@ifnotmtargexp{#2}{%
3179 \hypersetup{#1={#2}}%
3180 \def\next##1!,{}%
3181 }%
3182 \fi
3183 \fi
3184 \next
3185 }
3186 % \end{macrocode}
3187 % \end{macro}
3188 %
3189 % \begin{macro}{\hyxmp@auto@assign@data}
3190 % If certain metadata are unspecified, try to specify meaningful values
3191 % using data provided by author via other means (e.g.,~\cs{title} for
3192 % the document's title).
3193 % \begin{macrocode}
3194 \newcommand*{\hyxmp@auto@assign@data}{%
3195 % \end{macrocode}
3196 % If \cs{@pdflang} is not set, see if we can detect the document
3197 % language via either the \pkgname{babel} or \pkgname{polyglossia}
3198 % packages.
3199 % \begin{macrocode}
3200 \@if@def@and@nonempty{@pdflang}{%
3201 \let\hyxmp@dc@lang=\@pdflang
3202 }{%
3203 \hyxmp@detect@langs
3204 }%
3205 % \end{macrocode}
3206 % Replace an empty \cs{@pdfmetalang}. If \cs{@pdflang} is defined, use
3207 % that as the metadata language. Otherwise, use \xmpterm{x-default}.
3208 % \begin{macrocode}
3209 \ifx\@pdfmetalang\@empty
3210 \ifx\@pdflang\@empty
3211 \let\@pdfmetalang=\hyxmp@x@default
3212 \else
3213 \edef\@pdfmetalang{\@pdflang}%
3214 \fi
3215 \fi
3216 % \end{macrocode}
3217 % Identify various author-provided information that can be co-opted
3218 % for use as \acro{XMP} metadata.
3219 % \changes{v2.7}{2016/02/01}{Automatically use \cs{title} and \cs{author}
3220 % if \protect\optname{pdftitle} and \protect\optname{pdfauthor} are left
3221 % unspecified. Thanks to Maciej Radziejewski for the suggestion}
3222 % \changes{v5.3}{2020/05/29}{Consider other author-provided sources of
3223 % metadata. Thanks to Robin Schwab for proposing that
3224 % \protect\pkgname{hyperxmp} use the \Koma\ letter classes's metadata}
3225 % \changes{v5.5}{2020/07/24}{Moved the language-detection and
3226 % \protect\XeTeX\ date-detection code here from the
3227 % \protect\cs{hyxmp@at@end} block}
3228 % \changes{v5.5}{2020/07/24}{Use \protect\LuaTeX\ mechanisms, when available,
3229 % to automatically compute the page count}
3230 % \changes{v5.5}{2020/07/24}{Moved title and author autodetection to the
3231 % \protect\cs{AtEndPreamble}}
3232 % \begin{macrocode}
3233 \hyxmp@use@first@valid{pdfcontactemail}{\@pdfcontactemail}{%
3234 \scr@fromemail@var
3235 }%
3236 \hyxmp@set@koma@phones
3237 \hyxmp@use@first@valid{pdfcontactphone}{\@pdfcontactphone}{%
3238 \hyxmp@koma@phones
3239 }%
3240 \hyxmp@use@first@valid{pdfcontacturl}{\@pdfcontacturl}{%
3241 \scr@fromurl@var
3242 }%
3243 \hyxmp@use@first@valid{pdfsubtitle}{\@pdfsubtitle}{%
3244 \@subtitle
3245 }%
3246 \hyxmp@use@first@valid{pdfpublisher}{\@pdfpublisher}{%
3247 \@publishers
3248 }%
3249 % \end{macrocode}
3250 % We handle the \clsname{acmart} class specially. \clsname{acmart}
3251 % stores author-provided contact information in a structured format that
3252 % we can process fairly easily. Note that if the author is not using
3253 % the \clsname{acmart} class, \cs{hyxmp@parse@acmart} will have been
3254 % redefined to do nothing.
3255 % \begin{macrocode}
3256 \hyxmp@parse@acmart
3257 % \end{macrocode}
3258 % Most \acro{PDF} standards dictate that if the same metadata appear in
3259 % both the \acro{XMP} packet and the \acro{PDF} \pdfterm{Info}
3260 % dictionary, the metadata must match. This requirement poses a problem
3261 % for a user-unspecified \optname{pdfcreationdate} in the context of
3262 % \XeLaTeX\@. In this case we explicitly define |\@pdfcreationdate| as
3263 % |\hyxmp@today@pdf| to prevent the \cmdname{xdvipdfmx} back-end
3264 % processor from detecting a missing \pdfterm{CreationDate} in the
3265 % \pdfterm{Info} dictionary and adding its own---typically a few seconds
3266 % after \pkgname{hyperxmp} has constructed an \xmpprop{xmp:CreateDate}
3267 % for the \acro{XMP} metadata and leading to a metadata mismatch.
3268 % \begin{macrocode}
3269 \@ifundefined{XeTeXversion}{}{%
3270 \@ifmtargexp{\@pdfcreationdate}{%
3271 \let\@pdfcreationdate=\hyxmp@today@pdf
3272 }%
3273 {}%
3274 }%
3275 % \end{macrocode}
3276 % Query the document currently being built for page and byte counts.
3277 % \begin{macrocode}
3278 \hyxmp@query@self
3279 }
3280 % \end{macrocode}
3281 % \end{macro}
3282 %
3283 % Determine the size of the output file from the \emph{previous} run of
3284 % \LuaLaTeX\@. This action has to be performed before the
3285 % |\begin{document}| because at that point the size of the output file
3286 % is reset to zero. We use |\jobname.pdf| as the name of the output
3287 % file because |status.output_file_name| is not defined at this point.
3288 %
3289 % It's possible to use \pdfLaTeX's \cs{pdffilesize} primitive to query
3290 % the size of |\jobname.pdf| under \pdfLaTeX\@. Unfortunately, doing so
3291 % has a side effect of making \progname{latexmk} view the
3292 % \acro{PDF} file as an input file, which puts \progname{latexmk} in an
3293 % infinite build loop. (This was the case for \pkgname{hyperxmp} v5.5
3294 % and v5.6.) See the discussion at
3295 % \url{https://github.com/borisveytsman/acmart/issues/413} for more
3296 % information.
3297 % \changes{v5.7}{2020/11/01}{Do not automatically compute the
3298 % \protect\acro{PDF} file size under \protect\pdfLaTeX\ because this confuses
3299 % \protect\progname{latexmk}. Thanks to John Collins, Nelson Posse Lago,
3300 % Derek Dreyer, and the other contributors to
3301 % \protect\href{https://github.com/borisveytsman/acmart/issues/413}{\protect\pkgname{acmart} issue \#413},
3302 % ``Latexmk goes into an infinite loop even on sample files from ACM''}
3303 % \changes{v5.8}{2020/11/05}{Take
3304 % \protect\mbox{\protect\texttt{-\kern0pt-output-directory}} into
3305 % consideration when querying the output file size. Thanks to John
3306 % Collins for pointing out that the user can change the output
3307 % directory}
3308 % \begin{macrocode}
3309 \ifLuaTeX
3310 % \end{macrocode}
3311 % Now that we know we're running \LuaLaTeX\ we define a Lua function,
3312 % |get_pdf_size|, that takes the base name of the output file and
3313 % returns the number of bytes in the corresponding \acro{PDF} file. One
3314 % difficulty is that, at the time of this writing, \LuaTeX\ lacks a
3315 % mechanism for querying the full name of the output file. Our
3316 % workaround is a tad kludgy but seems to work. We walk the list of
3317 % command-line arguments for
3318 % \mbox{``|-\kern0pt-output-directory=|\meta{dir}''}. (We in fact
3319 % accept either one or two initial dashes and abbreviations as terse as
3320 % \mbox{``|-output-d|''}.) Then, we concatenate the output directory
3321 % (or ``|.|'' if unspecified), a path separator, the given base name of
3322 % the job, and a ``|.pdf|'' extension. Alas, different operating
3323 % systems use different path separators so we have to query the
3324 % operating-system type to select an appropriate separator: ``|\|'' on
3325 % Windows/\acro{DOS} and ``|/|'' on everything else.
3326 %
3327 % |get_pdf_size| is called regardless of whether we're producing
3328 % \acro{PDF} or \acro{DVI} output. We assume that even if the user
3329 % specified |--output-format=dvi|, the user's intention is eventually to
3330 % convert the document to \acro{PDF}.
3331 % \begin{macrocode}
3332 \begin{luacode*}
3333 function get_pdf_size (bname)
3334 % \end{macrocode}
3335 % Search the list of command-line arguments for the output directory.
3336 % \begin{macrocode}
3337 local odir = ""
3338 for _, opt in ipairs(arg) do
3339 local m = string.match(opt, "%-output%-d.-=(.*)")
3340 if m then
3341 odir = m
3342 end
3343 end
3344 % \end{macrocode}
3345 % Set the path separator to either ``|/|'' or ``|\|'', depending on the
3346 % operating system.
3347 % \begin{macrocode}
3348 local sep = "/"
3349 if os.type == "windows" or os.type == "msdos" then
3350 sep = "\\\\"
3351 end
3352 % \end{macrocode}
3353 % Concatenate the output directory, path separator, base name, and
3354 % \texttt{.pdf} extension. Do not insert a path separator if either
3355 % (1)~no output directory was specified, (2)~the output directory
3356 % already ends with the path separator, or (3)~the output directory ends
3357 % in a colon (and is therefore a relative directory) on
3358 % Windows/\acro{DOS}. As a few examples,
3359 %
3360 % \bigskip
3361 % \noindent
3362 % \begingroup
3363 % \let\bs=\textbackslash
3364 % \begin{tabular}{@{\quad\textbullet~}l@{~+~}l@{~+~}l@{~+~``\texttt{.pdf}''~=~}l@{}}
3365 % ``'' & ``\texttt{/}'' & ``\texttt{myfile}''
3366 % & ``\texttt{myfile.pdf}'' \\
3367 % ``\texttt{/docs}'' & ``\texttt{/}'' & ``\texttt{myfile}''
3368 % & ``\texttt{/docs/myfile.pdf}'' \\
3369 % ``\texttt{/docs/}'' & ``\texttt{/}'' & ``\texttt{myfile}''
3370 % & ``\texttt{/docs/myfile.pdf}'' \\
3371 % ``\texttt{C:\bs docs}'' & ``\texttt{\bs}'' & ``\texttt{myfile}''
3372 % & ``\texttt{C:\bs docs\bs myfile.pdf}'' \\
3373 % ``\texttt{C:\bs docs\bs}'' & ``\texttt{\bs}'' & ``\texttt{myfile}''
3374 % & ``\texttt{C:\bs docs\bs myfile.pdf}'' \\
3375 % ``\texttt{C:\bs}'' & ``\texttt{\bs}'' & ``\texttt{myfile}''
3376 % & ``\texttt{C:\bs myfile.pdf}'' \\
3377 % ``\texttt{C:}'' & ``\texttt{\bs}'' & ``\texttt{myfile}''
3378 % & ``\texttt{C:myfile.pdf}'' \\
3379 % \end{tabular}
3380 % \endgroup
3381 % \bigskip
3382 % \begin{macrocode}
3383 local dlast = string.sub(odir, -1)
3384 if odir == "" or dlast == sep or (dlast == ":" and sep == "\\\\") then
3385 sep = ""
3386 end
3387 local fname = odir .. sep .. bname .. ".pdf"
3388 % \end{macrocode}
3389 % Query the file size and return it.
3390 % \begin{macrocode}
3391 local nbytes = lfs.attributes(fname, "size")
3392 return nbytes
3393 end
3394 \end{luacode*}
3395 % \end{macrocode}
3396 % Now that we've defined |get_pdf_size| we invoke it, passing it
3397 % \cs{hyxmp@jobname} as the base name of the job. (Recall that
3398 % \cs{hyxmp@jobname} is the same as \cs{jobname} but with any
3399 % surrounding double quotes removed.) We store |get_pdf_size|'s
3400 % output---which will be empty if the \acro{PDF} file doesn't yet
3401 % exist---in \cs{hyxmp@prev@pdf@size}.
3402 % \begin{macrocode}
3403 \xdef\hyxmp@prev@pdf@size{%
3404 \luadirect{
3405 nbytes = get_pdf_size("\hyxmp@jobname")
3406 if nbytes then
3407 tex.write(nbytes)
3408 end
3409 }%
3410 }%
3411 \fi
3412 % \end{macrocode}
3413 %
3414 % \begin{macro}{\hyxmp@query@self}
3415 % Query the document currently being built to acquire page and byte counts.
3416 % \changes{v5.5}{2020/07/25}{Added this macro}
3417 % \begin{macrocode}
3418 \newcommand*{\hyxmp@query@self}{%
3419 % \end{macrocode}
3420 % \LaTeX's |totalpages| counter tracks the number of pages written. We
3421 % use this mechanism to assign \cs{@pdfnumpages}.
3422 % \changes{v5.11}{2022/10/16}{Use \protect\cs{thetotalpages} to compute
3423 % the page count in an engine-independent manner. Thanks to Ulrike
3424 % Fischer for recommending this mechanism}
3425 % \begin{macrocode}
3426 \@if@def@and@nonempty{@pdfnumpages}{%
3427 }{%
3428 \xdef\@pdfnumpages{\thetotalpages}%
3429 }%
3430 % \end{macrocode}
3431 % If \optname{pdfbytes} hasn't been set, set it to the output file's size
3432 % from the previous run.
3433 % \begin{macrocode}
3434 \hyxmp@use@first@valid{pdfbytes}{\@pdfbytes}{%
3435 \hyxmp@prev@pdf@size
3436 }%
3437 }
3438 % \end{macrocode}
3439 % \end{macro}
3440 %
3441 % \begin{macro}{\hyxmp@parse@acmart}
3442 % The \clsname{acmart} class stores a rich set of author metadata in its
3443 % \cs{addresses} macro. \cs{hyxmp@parse@acmart} extracts the contact
3444 % information for the first author and converts that to \acro{XMP}
3445 % metadata.
3446 % \changes{v5.3}{2020/05/30}{Added this macro}
3447 % \begin{macrocode}
3448 \newcommand*{\hyxmp@parse@acmart}{%
3449 \begingroup
3450 % \end{macrocode}
3451 % \begin{macro}{\@author}
3452 % \clsname{acmart} has already invoked
3453 % |\hypersetup{pdfauthor=|\dots|}| to specify the complete list of
3454 % authors. At this point, |\@author| is defined to produce a warning
3455 % message. We locally redefine it to do nothing.
3456 % \begin{macrocode}
3457 \let\@author=\@gobble
3458 % \end{macrocode}
3459 % \end{macro}
3460 % \begin{macro}{\email}
3461 % \begin{macro}{\hyxmp@address@val}
3462 % Within \cs{addresses}, \cs{email} is defined to accept two arguments,
3463 % the second of which is the author's email address.
3464 % \begin{macrocode}
3465 \def\email##1##2{%
3466 \def\hyxmp@address@val{##2}%
3467 \hyxmp@use@first@valid{pdfcontactemail}{\@pdfcontactemail}{%
3468 \hyxmp@address@val
3469 }%
3470 }%
3471 % \end{macrocode}
3472 % \end{macro}
3473 % \end{macro}
3474 % \begin{macro}{\streetaddress}
3475 % \begin{macro}{\hyxmp@address@val}
3476 % \cs{streetaddress} wraps the author's street address.
3477 % \begin{macrocode}
3478 \def\streetaddress##1{%
3479 \def\hyxmp@address@val{##1}%
3480 \hyxmp@use@first@valid{pdfcontactaddress}{\@pdfcontactaddress}{%
3481 \hyxmp@address@val
3482 }%
3483 }%
3484 % \end{macrocode}
3485 % \end{macro}
3486 % \end{macro}
3487 % \begin{macro}{\city}
3488 % \begin{macro}{\hyxmp@address@val}
3489 % \cs{city} wraps the author's city name.
3490 % \begin{macrocode}
3491 \def\city##1{%
3492 \def\hyxmp@address@val{##1}%
3493 \hyxmp@use@first@valid{pdfcontactcity}{\@pdfcontactcity}{%
3494 \hyxmp@address@val
3495 }%
3496 }%
3497 % \end{macrocode}
3498 % \end{macro}
3499 % \end{macro}
3500 % \begin{macro}{\state}
3501 % \begin{macro}{\hyxmp@address@val}
3502 % \cs{state} wraps the author's state or region name.
3503 % \begin{macrocode}
3504 \def\state##1{%
3505 \def\hyxmp@address@val{##1}%
3506 \hyxmp@use@first@valid{pdfcontactregion}{\@pdfcontactregion}{%
3507 \hyxmp@address@val
3508 }%
3509 }%
3510 % \end{macrocode}
3511 % \end{macro}
3512 % \end{macro}
3513 % \begin{macro}{\country}
3514 % \begin{macro}{\hyxmp@address@val}
3515 % \cs{country} wraps the author's country name.
3516 % \begin{macrocode}
3517 \def\country##1{%
3518 \def\hyxmp@address@val{##1}%
3519 \hyxmp@use@first@valid{pdfcontactcountry}{\@pdfcontactcountry}{%
3520 \hyxmp@address@val
3521 }%
3522 }%
3523 % \end{macrocode}
3524 % \end{macro}
3525 % \end{macro}
3526 % \begin{macro}{\postcode}
3527 % \begin{macro}{\hyxmp@address@val}
3528 % \cs{postcode} wraps the author's postal code.
3529 % \begin{macrocode}
3530 \def\postcode##1{%
3531 \def\hyxmp@address@val{##1}%
3532 \hyxmp@use@first@valid{pdfcontactpostcode}{\@pdfcontactpostcode}{%
3533 \hyxmp@address@val
3534 }%
3535 }%
3536 % \end{macrocode}
3537 % \end{macro}
3538 % \end{macro}
3539 % \begin{macro}{\affiliation}
3540 % We want to produce \acro{XMP} metadata for only a single affiliation.
3541 % Although \cs{hyxmp@use@first@valid} will ensure that only the first
3542 % email, city, country, etc.\ encountered is considered, we run the
3543 % first of one affiliation defining, say, a city and state but no
3544 % country and a subsequent affiliation defining a country. In that
3545 % case, the \acro{XMP} would include the first author's city and state
3546 % and the subsequent author's country. Hence, we define
3547 % \cs{affiliation} to ``self destruct'' after its first use, discarding
3548 % all further affiliations.
3549 % \begin{macrocode}
3550 \def\affiliation##1##2{%
3551 ##2%
3552 \let\affiliation=\@gobbletwo
3553 }%
3554 % \end{macrocode}
3555 % \end{macro}
3556 %
3557 % We want to evaluate \cs{addresses} with the preceding local
3558 % definitions in effect, but we don't want to typeset any text appearing
3559 % in the string. Hence, we ``typeset'' \cs{addresses} within a box that
3560 % is subsequently discarded.
3561 % \begin{macrocode}
3562 \setbox0=\hbox{\addresses}%
3563 \endgroup
3564 % \end{macrocode}
3565 %
3566 % \clsname{acmart} supports other relevant metadata in addition to the
3567 % authors' mailing addresses. For instance, papers accepted for
3568 % publication indicate their \acro{DOI} number. However, papers under
3569 % review will contain either a placeholder \acro{DOI},
3570 % ``10.1145/nnnnnnn.nnnnnnn'', or the example \acro{DOI} specified in
3571 % the \clsname{acmart} example document, ``10.1145/1122445.1122456''.
3572 % We ignore both of those \acro{DOI}s.
3573 % \changes{v5.4}{2020/06/16}{Bug fix: Correct a missing ``else'' argument
3574 % in two invocations of \protect\cs{@if@def@and@nonempty}}
3575 % \begin{macrocode}
3576 \@if@def@and@nonempty{@acmDOI}{%
3577 \IfSubStr{\@acmDOI}{10.1145/1122445.1122456}{}{%
3578 \IfSubStr{\@acmDOI}{10.1145/nnnnnnn.nnnnnnn}{}{%
3579 \hyxmp@use@first@valid{pdfdoi}{\@pdfdoi}{%
3580 \@acmDOI
3581 }%
3582 }%
3583 }%
3584 }%
3585 {}%
3586 % \end{macrocode}
3587 % \begin{macro}{\hyxmp@strip@isbn@date}
3588 % \begin{macro}{\hyxmp@acm@isbn}
3589 % Papers appearing in conference proceedings specify the proceedings'
3590 % \acro{ISBN}. As with \cs{@acmDOI} above, we ignore both the
3591 % placeholder \acro{ISBN}, ``978-x-xxxx-xxxx-x/YY/MM'', and the example
3592 % \acro{ISBN}, ``978-1-4503-XXXX-X/18/06''. We also strip off the
3593 % ``\texttt{/}\meta{year}\texttt{/}\meta{month}'' suffix so as to
3594 % include a true \acro{ISBN} in the \acro{XMP} metadata.
3595 % \begin{macrocode}
3596 \@if@def@and@nonempty{@acmISBN}{%
3597 \IfSubStr{\@acmISBN}{XXXX}{}{%
3598 \IfSubStr{\@acmISBN}{xxxx}{}{%
3599 \def\hyxmp@strip@isbn@date##1/##2!{##1}%
3600 \edef\hyxmp@acm@isbn{%
3601 \expandafter\hyxmp@strip@isbn@date\@acmISBN/!%
3602 }%
3603 \hyxmp@use@first@valid{pdfisbn}{\@pdfisbn}{%
3604 \hyxmp@acm@isbn
3605 }%
3606 }%
3607 }%
3608 }%
3609 {}%
3610 % \end{macrocode}
3611 % \end{macro}
3612 % \end{macro}
3613 % \begin{macro}{\hyxmp@acm@publisher}
3614 % The publisher is of course \acro{ACM}.
3615 % \begin{macrocode}
3616 \def\hyxmp@acm@publisher{Association for Computing Machinery}%
3617 \hyxmp@use@first@valid{pdfpublisher}{\@pdfpublisher}{%
3618 \hyxmp@acm@publisher
3619 }%
3620 % \end{macrocode}
3621 % Use the journal name if defined, otherwise the book name (for
3622 % conference proceedings).
3623 % \begin{macrocode}
3624 \hyxmp@use@first@valid{pdfpublication}{\@pdfpublication}{%
3625 \@journalName,%
3626 \@acmBooktitle,%
3627 \@acmConference
3628 }%
3629 % \end{macrocode}
3630 % \end{macro}
3631 % \begin{macro}{\hyxmp@acm@pubtype}
3632 % \clsname{acmart} makes clear whether it's typesetting a journal
3633 % article. If it's not a journal, we assume it's a book (conference
3634 % proceedings).
3635 % \begin{macrocode}
3636 \if@ACM@journal
3637 \def\hyxmp@acm@pubtype{journal}%
3638 \else
3639 \def\hyxmp@acm@pubtype{book}%
3640 \fi
3641 \hyxmp@use@first@valid{pdfpubtype}{\@pdfpubtype}{%
3642 \hyxmp@acm@pubtype
3643 }%
3644 % \end{macrocode}
3645 % \end{macro}
3646 % Journal articles have a volume and issue number.
3647 % \begin{macrocode}
3648 \hyxmp@use@first@valid{pdfvolumenum}{\@pdfvolumenum}{%
3649 \@acmVolume
3650 }%
3651 \hyxmp@use@first@valid{pdfissuenum}{\@pdfissuenum}{%
3652 \@acmNumber
3653 }%
3654 }
3655 % \end{macrocode}
3656 % \end{macro}
3657 %
3658 % Nullify \cs{hyxmp@parse@acmart} if the author is not using the
3659 % \clsname{acmart} class.
3660 % \begin{macrocode}
3661 \@ifclassloaded{acmart}{}{\let\hyxmp@parse@acmart=\relax}
3662 % \end{macrocode}
3663 %
3664 % \begin{macro}{\hyxmp@dc@lang}
3665 % \cs{hyxmp@dc@lang} is a comma-separated list of all languages used in
3666 % the document.
3667 % \begin{macrocode}
3668 \let\hyxmp@dc@lang=\@empty
3669 % \end{macrocode}
3670 % \end{macro}
3671 %
3672 % \begin{macro}{\hyxmp@detect@langs}
3673 % If \optname{pdflang} was not specified, try to determine the document
3674 % language(s) using either \pkgname{babel}'s or \pkgname{polyglossia}'s
3675 % definitions.
3676 % \changes{v5.3}{2020/06/08}{Acquire the default language from the
3677 % \protect\pkgname{polyglossia} package, if loaded. Thanks to
3678 % Robin Schwab for bringing that package to my attention}
3679 % \changes{v5.4}{2020/06/16}{Refactored language detection into a
3680 % separate command}
3681 % \changes{v5.5}{2020/07/21}{Set the language(s) immediately instead of
3682 % deferring them to \protect\cs{hyxmp@set@dc@lang}}
3683 % \begin{macrocode}
3684 \newcommand*{\hyxmp@detect@langs}{%
3685 \@ifundefined{mainbcp47id}{%
3686 \@ifundefined{LocaleForEach}{%
3687 % \end{macrocode}
3688 % The document doesn't appear to have loaded either \pkgname{babel} or
3689 % \pkgname{polyglossia}. In this case we have one small task to do. In
3690 % older versions of \pkgname{hyperref}, \cs{@pdflang} is set to
3691 % \cs{@empty} if \optname{pdflang} is not specified. In newer versions
3692 % of \pkgname{hyperref}, \cs{@pdflang} is set to \cs{relax} if
3693 % \optname{pdflang} is not specified. The latter is a bit problematic
3694 % for \pkgname{hyperxmp} because it makes \cs{@pdflang} non-expandable,
3695 % which causes a literal ``\cs{@pdflang}'' to be written as \acro{XMP}
3696 % metadata. To avoid that situation we explicitly set \cs{@pdflang} to
3697 % \cs{@empty} to avoid problems with non-expandable symbols.
3698 % \changes{v2.3a}{2013/04/16}{Bug fix: Redefine \cs{@pdflang} as
3699 % \cs{@empty} when \protect\pkgname{hyperref} has set
3700 % it to \cs{relax}}
3701 % \begin{macrocode}
3702 \let\@pdflang=\@empty
3703 }{%
3704 % \end{macrocode}
3705 % \changes{v5.4}{2020/06/17}{Added support for \protect\pkgname{babel}}
3706 % \changes{v5.5}{2020/07/16}{Store the main language in \protect\cs{@pdflang}.
3707 % Thanks to Javier Bezos for his help with the \protect\pkgname{hyperxmp}
3708 % code and for modifying \protect\pkgname{babel} for
3709 % \protect\pkgname{hyperxmp}'s benefit}
3710 % \begin{macro}{\hyxmp@dc@lang}
3711 % \begin{macro}{\hyxmp@lang@tag}
3712 % \begin{macro}{\hyxmp@lang@name}
3713 % \begin{macro}{\@pdflang}
3714 % Use \pkgname{babel}'s \cs{LocaleForEach} and \cs{getlocaleproperty} to
3715 % set \cs{@pdflang} to the document's main language and
3716 % \cs{hyxmp@dc@lang} to a comma-separated list of all languages used.
3717 % \begin{macrocode}
3718 \BabelEnsureInfo
3719 \LocaleForEach{%
3720 \getlocaleproperty\hyxmp@lang@tag{##1}{identification/tag.bcp47}%
3721 \ifx\hyxmp@dc@lang\@empty
3722 \xdef\hyxmp@dc@lang{\hyxmp@lang@tag}%
3723 \else
3724 \xdef\hyxmp@dc@lang{\hyxmp@dc@lang,\hyxmp@lang@tag}%
3725 \fi
3726 \def\hyxmp@lang@name{##1}%
3727 \ifx\hyxmp@lang@name\bbl@main@language
3728 \edef\@pdflang{\hyxmp@lang@tag}%
3729 \fi
3730 }%
3731 }%
3732 }{%
3733 % \end{macrocode}
3734 % \end{macro}
3735 % \end{macro}
3736 % \end{macro}
3737 % \end{macro}
3738 % Use \pkgname{polyglossia}'s \cs{mainbcp47id} as the document's main
3739 % language and its \cs{xpg@bcp@loaded} as a comma-separated list of all
3740 % document languages.
3741 % \begin{macrocode}
3742 \xdef\@pdflang{\csname mainbcp47id\endcsname}%
3743 \edef\hyxmp@dc@lang{\xpg@bcp@loaded}%
3744 }%
3745 }
3746 % \end{macrocode}
3747 % \end{macro}
3748 %
3749 %
3750 % \subsection{Manipulating author-supplied data}
3751 %
3752 % The author provides metadata information to \pkgname{hyperxmp} via
3753 % package options to \pkgname{hyperref} or via \pkgname{hyperref}'s
3754 % |\hypersetup| command. The functions in this section convert
3755 % author-supplied lists (e.g.,~|pdfkeywords={foo, bar, baz}|) into
3756 % \LaTeX\ lists (e.g.,~|\@elt {foo}| |\@elt {bar}| |\@elt {baz}|) that
3757 % can be more easily manipulated (Section~\ref{sec:list-manip}); parse
3758 % dates in both \acro{PDF} and \acro{XMP} formats
3759 % (Section~\ref{sec:date-manip}; trim spaces off the ends of strings
3760 % (Section~\ref{sec:trim-spaces}); convert text to \acro{XML}
3761 % (e.g.,~from |<[email protected]>| to
3762 % |&lt;[email protected]&gt;|) (Section~\ref{sec:text-xml});
3763 % simplify the pretty-printing of a begin tag, \acro{XML} text, and end
3764 % tag (Section~\ref{sec:output-xml}; and provide metadata in multiple
3765 % languages (Section~\ref{sec:lang-alt}).
3766 %
3767 % \subsubsection{List manipulation}
3768 % \label{sec:list-manip}
3769 %
3770 % We define a macro for converting a list of comma-separated elements
3771 % (e.g.,~the list of \acro{PDF} keywords) to a list of \LaTeX\
3772 % |\@elt|-separated elements.
3773 %
3774 % \begin{macro}{\hyxmp@commas@to@list}
3775 % Given a macro name~(|#1|) and a comma-separated list~(|#2|), define
3776 % the macro name as the elements of the list, each preceded by |\@elt|.
3777 % (Executing the macro therefore applies |\@elt| to each element in
3778 % turn.)
3779 % \begin{macrocode}
3780 \newcommand*{\hyxmp@commas@to@list}[2]{%
3781 \gdef#1{}%
3782 \expandafter\hyxmp@commas@to@list@i\expandafter#1#2,,%
3783 }
3784 % \end{macrocode}
3785 % \end{macro}
3786 %
3787 % \begin{macro}{\hyxmp@commas@to@list@i}
3788 % \begin{macro}{\next}
3789 % Recursively construct macro~|#1| from comma-separated list~|#2|. Stop
3790 % if |#2| is empty.
3791 % \begin{macrocode}
3792 \def\hyxmp@commas@to@list@i#1#2,{%
3793 \gdef\hyxmp@sublist{#2}%
3794 \ifx\hyxmp@sublist\@empty
3795 \let\next=\relax
3796 \else
3797 \hyxmp@trimspaces\hyxmp@sublist
3798 \@cons{#1}{{\hyxmp@sublist}}%
3799 \def\next{\hyxmp@commas@to@list@i{#1}}%
3800 \fi
3801 \next
3802 }
3803 % \end{macrocode}
3804 % \end{macro}
3805 % \end{macro}
3806 %
3807 % \begin{macro}{\xmpcomma}
3808 % Because \pkgname{hyperxmp} splits lists at commas, a comma cannot
3809 % normally be used within a list. We there provide an |\xmpcomma| macro
3810 % that can expand to either a true comma or a placeholder character
3811 % depending on the situation. Here, we bind it to a comma so it can be
3812 % used in \emph{any} \pkgname{hyperxmp} option, not just those that treat
3813 % commas specially.
3814 % \changes{v2.0}{2012/08/25}{Added this macro}
3815 % \changes{v2.2}{2012/12/07}{Changed the default from
3816 % \texttt{\string\string\string\relax} to an ordinary comma}
3817 % \begin{macrocode}
3818 \def\xmpcomma{,}%
3819 % \end{macrocode}
3820 % \end{macro}
3821 %
3822 % \begin{macro}{\hyxmp@comma}
3823 % This is what |\xmpcomma| maps to during list construction. We assume
3824 % that documents will never otherwise use an \acro{ETX} (|^^C|)
3825 % character in their \acro{XMP} metadata.
3826 % \changes{v2.0}{2012/08/25}{Added this macro}
3827 % \begin{macrocode}
3828 \bgroup
3829 \catcode`\^^C=11
3830 \gdef\hyxmp@comma{^^C}
3831 \egroup
3832 % \end{macrocode}
3833 % \end{macro}
3834 %
3835 % \begin{macro}{\hyxmp@uscore}
3836 % This is what |\_| temporarily maps to during packet construction.
3837 % Because underscores are replaced by spaces, we need a mechanism to
3838 % preserve user-specified underscores (e.g.,~in email addresses). We
3839 % assume that documents will never otherwise use an \acro{NAK} (|^^U|)
3840 % character in their \acro{XMP} metadata.
3841 % \changes{v2.5}{2014/06/19}{Added this macro}
3842 % \begin{macrocode}
3843 \bgroup
3844 \catcode`\^^U=11
3845 \gdef\hyxmp@uscore{^^U}
3846 \egroup
3847 % \end{macrocode}
3848 % \end{macro}
3849 %
3850 % \begin{macro}{\xmpquote}
3851 % Adobe Acrobat likes to see double quotes around list elements that
3852 % contain commas when the entire list appears within a single \acro{XMP}
3853 % tag (e.g.,~|<pdf:Keywords>|). However, it doesn't like to see double
3854 % quotes around list elements that contain commas when the list is
3855 % broken up into individual components (i.e.,~using |<rdf:li>| tags).
3856 % We therefore introduce an |\xmpquote| macro that quotes or doesn't
3857 % quote its argument based on context. Here, we bind |\xmpquote| to
3858 % |\relax| to prevent it from prematurely quoting or not quoting.
3859 % \changes{v2.0}{2012/09/10}{Added this macro}
3860 % \begin{macrocode}
3861 \let\xmpquote=\relax
3862 % \end{macrocode}
3863 % \end{macro}
3864 %
3865 % \begin{macro}{\xmptilde}
3866 % As a convenience for the user, we define |\xmptilde| as a category~12
3867 % (other) ``|~|'' character.
3868 % \changes{v2.4}{2014/01/01}{Added this macro}
3869 % \begin{macrocode}
3870 \bgroup
3871 \catcode`\~=12%
3872 \gdef\xmptilde{~}%
3873 \egroup
3874 % \end{macrocode}
3875 % \end{macro}
3876 %
3877 % \begin{macro}{\XMPTruncateList}
3878 % \changes{v2.0}{2012/09/07}{Added this macro}
3879 % \changes{v2.3b}{2013/07/18}{Made all definitions local to avoid
3880 % spurious \texttt{Too many unprocessed floats} errors when running
3881 % with \string\pkgname{memoir}}
3882 % \changes{v4.0}{2019/03/31}{Deprecated this macro}
3883 % \begin{macro}{\hyxmp@temp@str}
3884 % \begin{macro}{\hyxmp@temp@list}
3885 % \begin{macro}{\@elt}
3886 % As a workaround for the inability of older Adobe Acrobat versions to
3887 % display author lists correctly we introduce a hack that replaces a
3888 % list with its first element. One can then write
3889 % ``|\XMPTruncateList{pdfauthor}|'' and have Adobe Acrobat display the
3890 % author list correctly.
3891 % \begin{macrocode}
3892 \newcommand{\XMPTruncateList}[1]{{%
3893 \PackageWarning{hyperxmp}{%
3894 \noexpand\XMPTruncateList has been deprecated since\MessageBreak
3895 hyperxmp 4.0 and may be removed in future\MessageBreak
3896 versions of the package. \noexpand\XMPTruncateList\MessageBreak
3897 was found}%
3898 \edef\hyxmp@temp@str{\csname hyxmp@#1\endcsname}%
3899 \hyxmp@commas@to@list{\hyxmp@temp@list}{\hyxmp@temp@str}%
3900 \def\@elt##1{%
3901 \expandafter\gdef\csname @#1\endcsname{##1}%
3902 \let\@elt=\@gobble
3903 }
3904 \hyxmp@temp@list
3905 }}
3906 % \end{macrocode}
3907 % \end{macro}
3908 % \end{macro}
3909 % \end{macro}
3910 % \end{macro}
3911 %
3912 %
3913 % \subsubsection{Date manipulation}
3914 % \label{sec:date-manip}
3915 %
3916 % \pkgname{hyperxmp} needs to manipulate two types of date (really,
3917 % timestamp) formats: \acro{PDF} format and \acro{XMP} format.
3918 % \acro{PDF} timestamps are of the form
3919 % ``|D:|\textsc{yyyymmdd}hhmmss|+|\textsc{tt}|'|tt|'|''
3920 % \makeatletter
3921 % \@ifundefined{pdffeedback}{^^A
3922 % \@ifundefined{pdfcreationdate}{^^A
3923 % }{^^A
3924 % (e.g.,~\texttt{\pdfcreationdate})^^A
3925 % }^^A
3926 % }{^^A
3927 % (e.g.,~\texttt{\pdffeedback creationdate})^^A
3928 % }
3929 % \makeatother
3930 % ~\cite{Adobe2008:PDF}, while \acro{XMP} timestamps are of the form
3931 % ``\textsc{yyyy}|-|\textsc{mm}|-|\textsc{dd}|T|hh|:|mm|:|ss|+|\textsc{tt}|:|tt''
3932 % \makeatletter
3933 % \@ifundefined{pdffeedback}{^^A
3934 % \@ifundefined{pdfcreationdate}{^^A
3935 % \unskip
3936 % }{^^A
3937 % (e.g.,~\texttt{\expandafter\hyxmp@pdf@to@xmp@date\pdfcreationdate})^^A
3938 % }^^A
3939 % }{^^A
3940 % (e.g.,~\texttt{\expandafter\hyxmp@pdf@to@xmp@date\pdffeedback creationdate})^^A
3941 % }^^A
3942 % \makeatother
3943 % ~\cite{Adobe2012:XMP}. The |\hyxmp@as@pdf@date| and
3944 % |\hyxmp@as@xmp@date| macros defined in this section facilitate
3945 % timestamp conversions to \acro{PDF} and \acro{XMP} formats,
3946 % respectively.
3947 %
3948 % \begin{macro}{\hyxmp@first@char}
3949 % \begin{macro}{\hyxmp@first@char@i}
3950 % Return the first character of a string. This macro is fully expandable.
3951 % \begin{macrocode}
3952 \def\hyxmp@first@char#1{\hyxmp@first@char@i#1\relax}
3953 \def\hyxmp@first@char@i#1#2\relax{#1}
3954 % \end{macrocode}
3955 % \end{macro}
3956 % \end{macro}
3957 %
3958 % \begin{macro}{\hyxmp@as@xmp@date}
3959 % If necessary, convert a timestamp to \acro{XMP} format. That is, if
3960 % the timestamp is in \acro{PDF} format, convert it; otherwise, leave it
3961 % unmodified. This macro is fully expandable.
3962 % \changes{v3.2}{2017/02/20}{Added this macro}
3963 % \begin{macrocode}
3964 \def\hyxmp@as@xmp@date#1{%
3965 \expandafter\ifnum\expandafter`\hyxmp@first@char@i#1\relax=`D
3966 \hyxmp@pdf@to@xmp@date{#1}%
3967 \else
3968 #1%
3969 \fi
3970 }
3971 % \end{macrocode}
3972 % \end{macro}
3973 %
3974 % \begin{macro}{\hyxmp@pdf@to@xmp@date}
3975 % Convert a timestamp from \acro{PDF} format to \acro{XMP} format. This
3976 % macro is fully expandable.
3977 % \changes{v2.4}{2013/12/24}{Added this macro}
3978 % \begin{macrocode}
3979 \def\hyxmp@pdf@to@xmp@date#1:#2#3#4#5#6#7#8#9{%
3980 #2#3#4#5-#6#7-#8#9%
3981 \hyxmp@parse@time
3982 }
3983 % \end{macrocode}
3984 % \end{macro}
3985 %
3986 % \begin{macro}{\hyxmp@parse@time}
3987 % This is a helper function for |\hyxmp@pdf@to@xmp@date|.
3988 % |\hyxmp@pdf@to@xmp@date| proper parses only the year, month, and day
3989 % then calls |\hyxmp@parse@time|. |\hyxmp@parse@time| parses the
3990 % hours, minutes, and seconds then calls |\hyxmp@parse@tz@char|.
3991 % \changes{v2.4}{2013/12/24}{Added this macro}
3992 % \begin{macrocode}
3993 \def\hyxmp@parse@time#1#2#3#4#5#6{%
3994 T#1#2:#3#4:#5#6%
3995 \hyxmp@parse@tz@char
3996 }
3997 % \end{macrocode}
3998 % \end{macro}
3999 %
4000 % \begin{macro}{\hyxmp@parse@tz@char}
4001 % This is another helper function for |\hyxmp@pdf@to@xmp@date|. So far,
4002 % the date and time have been parsed. |\hyxmp@parse@tz@char| parses the
4003 % first character of the timezone descriptor. This can be one of
4004 % ``|+|'' for eastern timezones (\mbox{UTC$+x$}, including Asia,
4005 % Oceania, and most of Europe), ``|-|'' for western timezones
4006 % (\mbox{UTC$-x$}, primarily the Americas), or ``|Z|'' for Zulu time
4007 % (\mbox{UTC$+0$}). Timezones beginning with ``|+|'' or ``|-|'' are
4008 % followed by an offset in hours and minutes (parsed by
4009 % |\hyxmp@parse@tz|; timezones beginning with ``|Z|'' are not.
4010 % \changes{v2.4}{2013/12/24}{Added this macro}
4011 % \begin{macrocode}
4012 \def\hyxmp@parse@tz@char#1{%
4013 #1%
4014 \ifx#1-%
4015 \expandafter\hyxmp@parse@tz
4016 \else
4017 \ifx#1+%
4018 \expandafter\hyxmp@parse@tz
4019 \fi
4020 \fi
4021 }
4022 % \end{macrocode}
4023 % \end{macro}
4024 %
4025 % \begin{macro}{\hyxmp@parse@tz}
4026 % This is the final helper function for |\hyxmp@pdf@to@xmp@date|. It
4027 % parses the piece of the timezone comprising the offset from
4028 % Coordinated Universal Time, measured in hours and minutes.
4029 % \changes{v2.4}{2013/12/24}{Added this macro}
4030 % \begin{macrocode}
4031 \def\hyxmp@parse@tz#1'#2'{%
4032 #1:#2%
4033 }
4034 % \end{macrocode}
4035 % \end{macro}
4036 %
4037 % \begin{macro}{\hyxmp@as@pdf@date}
4038 % If necessary, convert a timestamp to \acro{PDF} format. That is, if
4039 % the timestamp is in \acro{XMP} format, convert it; otherwise, leave it
4040 % unmodified. This macro is fully expandable.
4041 % \changes{v3.2}{2017/02/20}{Added this macro}
4042 % \begin{macrocode}
4043 \def\hyxmp@as@pdf@date#1{%
4044 \expandafter\ifx\hyxmp@first@char@i#1\relax D%
4045 #1%
4046 \else
4047 \hyxmp@xmp@to@pdf@date{#1}%
4048 \fi
4049 }
4050 % \end{macrocode}
4051 % \end{macro}
4052 %
4053 % \begin{macro}{\hyxmp@xmp@to@pdf@date}
4054 % Convert a timestamp from \acro{XMP} format to \acro{PDF} format. This
4055 % macro is fully expandable.
4056 % \changes{v2.4}{2013/12/24}{Added this macro}
4057 % \begin{macrocode}
4058 \def\hyxmp@xmp@to@pdf@date#1{%
4059 D:\hyxmp@xmp@to@pdf@date@i#1\relax\relax
4060 }
4061 % \end{macrocode}
4062 % \end{macro}
4063 %
4064 % \begin{macro}{\hyxmp@xmp@to@pdf@date@i}
4065 % Parse the year for |\hyxmp@xmp@to@pdf@date|.
4066 % \begin{macrocode}
4067 \def\hyxmp@xmp@to@pdf@date@i#1#2#3#4#5#6{%
4068 #1#2#3#4%
4069 \ifx#5-%
4070 \expandafter\hyxmp@xmp@to@pdf@date@ii\expandafter#6%
4071 \fi
4072 }
4073 % \end{macrocode}
4074 % \end{macro}
4075 %
4076 % \begin{macro}{\hyxmp@xmp@to@pdf@date@ii}
4077 % Parse the month for |\hyxmp@xmp@to@pdf@date|.
4078 % \begin{macrocode}
4079 \def\hyxmp@xmp@to@pdf@date@ii#1#2#3#4{%
4080 #1#2%
4081 \ifx#3-%
4082 \expandafter\hyxmp@xmp@to@pdf@date@iii\expandafter#4%
4083 \fi
4084 }
4085 % \end{macrocode}
4086 % \end{macro}
4087 %
4088 % \begin{macro}{\hyxmp@xmp@to@pdf@date@iii}
4089 % Parse the day for |\hyxmp@xmp@to@pdf@date|.
4090 % \begin{macrocode}
4091 \def\hyxmp@xmp@to@pdf@date@iii#1#2#3#4{%
4092 #1#2%
4093 \ifx#3T%
4094 \expandafter\hyxmp@xmp@to@pdf@date@iv\expandafter#4%
4095 \fi
4096 }
4097 % \end{macrocode}
4098 % \end{macro}
4099 %
4100 % \begin{macro}{\hyxmp@xmp@to@pdf@date@iv}
4101 % Parse the hour for |\hyxmp@xmp@to@pdf@date|.
4102 % \begin{macrocode}
4103 \def\hyxmp@xmp@to@pdf@date@iv#1#2#3#4{%
4104 #1#2%
4105 \ifx#3:%
4106 \expandafter\hyxmp@xmp@to@pdf@date@v\expandafter#4%
4107 \fi
4108 }
4109 % \end{macrocode}
4110 % \end{macro}
4111 %
4112 % \begin{macro}{\hyxmp@xmp@to@pdf@date@v}
4113 % Parse the minute for |\hyxmp@xmp@to@pdf@date|.
4114 % \begin{macrocode}
4115 \def\hyxmp@xmp@to@pdf@date@v#1#2#3#4{%
4116 #1#2%
4117 \ifx#3:%
4118 \expandafter\hyxmp@xmp@to@pdf@date@vi\expandafter#4%
4119 \fi
4120 }
4121 % \end{macrocode}
4122 % \end{macro}
4123 %
4124 % \begin{macro}{\hyxmp@gobbletwo}
4125 % This is exactly the same as \LaTeXe's |\@gobbletwo| but needs to be a
4126 % different literal for |\hyxmp@xmp@to@pdf@date@vii|'s pattern-matching
4127 % to work.
4128 % \begin{macrocode}
4129 \let\hyxmp@gobbletwo=\@gobbletwo
4130 % \end{macrocode}
4131 % \end{macro}
4132 %
4133 % \begin{macro}{\hyxmp@xmp@to@pdf@date@vi}
4134 % Parse the second for |\hyxmp@xmp@to@pdf@date|. The challenge here is
4135 % that we need to handle four cases for the character following the
4136 % seconds---``|+|'', ``|-|'', ``|Z|'', and no character---without sacrificing
4137 % expandability. Our tricky solution is to insert a |\@gobbletwo| as a
4138 % sentinel and let |\hyxmp@xmp@to@pdf@date@vi| discard everything up to
4139 % that sentinel (i.e.,~all the other conditionals).
4140 % \begin{macrocode}
4141 \def\hyxmp@xmp@to@pdf@date@vi#1#2#3#4{%
4142 #1#2%
4143 \ifx#3+%
4144 +\expandafter\hyxmp@xmp@to@pdf@date@vii
4145 \fi
4146 \ifx#3-%
4147 -\expandafter\hyxmp@xmp@to@pdf@date@vii
4148 \fi
4149 \ifx#3Z%
4150 Z%
4151 \fi
4152 \ifx#3\relax
4153 \expandafter\hyxmp@gobbletwo
4154 \fi
4155 \@gobbletwo #4%
4156 }
4157 % \end{macrocode}
4158 % \end{macro}
4159 %
4160 % \begin{macro}{\hyxmp@xmp@to@pdf@date@vii}
4161 % Parse the time-zone hours for |\hyxmp@xmp@to@pdf@date|.
4162 % \begin{macrocode}
4163 \def\hyxmp@xmp@to@pdf@date@vii#1\@gobbletwo#2#3#4#5{%
4164 #2#3%
4165 \ifx#4:%
4166 \expandafter\hyxmp@xmp@to@pdf@date@viii\expandafter#5%
4167 \fi
4168 }
4169 % \end{macrocode}
4170 % \end{macro}
4171 %
4172 % \begin{macro}{\hyxmp@xmp@to@pdf@date@viii}
4173 % Parse the time-zone minutes for |\hyxmp@xmp@to@pdf@date|.
4174 % \begin{macrocode}
4175 \def\hyxmp@xmp@to@pdf@date@viii#1#2#3#4{%
4176 '#1#2'%
4177 }
4178 % \end{macrocode}
4179 % \end{macro}
4180 %
4181 % \begin{macro}{\hyxmp@today@xmp@define}
4182 % Use \tex\ primitives to define a given macro as today's date in
4183 % \textsc{yyyy}-\textsc{mm}-\textsc{dd}|T|hh|:|mm|Z| format.
4184 % \changes{v2.4}{2013/12/21}{Added this macro}
4185 % \changes{v3.0}{2016/07/03}{Modified to accept the name of a macro
4186 % to define}
4187 % \changes{v3.2}{2017/02/21}{Modified to include hours and minutes}
4188 % \changes{v5.0}{2020/03/16}{Modified to specify UTC}
4189 % \begin{macrocode}
4190 \def\hyxmp@today@xmp@define#1{%
4191 % \end{macrocode}
4192 % The date is a straightforward representation of \tex's |\year|,
4193 % |\month|, and |\day| primitives, with the latter two zero-padded to
4194 % two digits apiece.
4195 % \begin{macrocode}
4196 \xdef#1{\the\year}%
4197 \ifnum\month<10
4198 \xdef#1{#1-0\the\month}%
4199 \else
4200 \xdef#1{#1-\the\month}%
4201 \fi
4202 \ifnum\day<10
4203 \xdef#1{#1-0\the\day}%
4204 \else
4205 \xdef#1{#1-\the\day}%
4206 \fi
4207 % \end{macrocode}
4208 % \tex\ does not provide the time in terms of separate hours and minutes
4209 % but rather as the total number of minutes since midnight (|\time|).
4210 % There's no mechanism in \tex\ to query the number of seconds since
4211 % midnight or the timezone so we omit those fields when defining
4212 % macro~|#1|.
4213 % \begin{macrocode}
4214 \@hyxmp@count=\time
4215 \divide\@hyxmp@count by 60
4216 \ifnum\@hyxmp@count<10
4217 \xdef#1{#1T0\the\@hyxmp@count}%
4218 \else
4219 \xdef#1{#1T\the\@hyxmp@count}%
4220 \fi
4221 \multiply\@hyxmp@count by -60
4222 \advance\@hyxmp@count by \time
4223 \ifnum\@hyxmp@count<10
4224 \xdef#1{#1:0\the\@hyxmp@count}%
4225 \else
4226 \xdef#1{#1:\the\@hyxmp@count}%
4227 \fi
4228 \xdef#1{#1Z}%
4229 }
4230 % \end{macrocode}
4231 % \end{macro}
4232 %
4233 % \begin{macro}{\hyxmp@try@today}
4234 % If |\hyxmp@today@xmp| is still empty and |#1| is defined, evaluate |#2|.
4235 % Otherwise, do nothing.
4236 % \begin{macrocode}
4237 \def\hyxmp@try@today#1#2{%
4238 \@ifmtargexp{\hyxmp@today@xmp}{%
4239 \@ifundefined{#1}{}{#2}%
4240 }%
4241 {}%
4242 }
4243 % \end{macrocode}
4244 % \end{macro}
4245 %
4246 % \begin{macro}{\hyxmp@today@xmp}
4247 % Define |\hyxmp@today@xmp| as the current date and (if available) time and
4248 % timezone in \acro{XMP} \xmpterm{Date} format~\cite{Adobe2012:XMP}.
4249 % \changes{v2.4}{2013/12/24}{Modified the code to parse the time and
4250 % timezone from \texttt{\string\string\string\pdfcreationdate}, as
4251 % proposed by Florian Breitwieser}
4252 % \changes{v5.0}{2020/03/15}{Support \string\XeTeX's \string\cs{filemoddate}}
4253 % \begin{macrocode}
4254 \def\hyxmp@today@xmp{}
4255 % \end{macrocode}
4256 % Case 1: |\pdfcreationdate| is defined (\pdfLaTeX\ and pre-0.85 \LuaLaTeX).
4257 % \begin{macrocode}
4258 \hyxmp@try@today{pdfcreationdate}{%
4259 \edef\hyxmp@today@xmp{\expandafter\hyxmp@pdf@to@xmp@date\pdfcreationdate}%
4260 }
4261 % \end{macrocode}
4262 % Case 2: |\pdffeedback| is defined (\LuaLaTeX~0.85+).
4263 % \begin{macrocode}
4264 \hyxmp@try@today{pdffeedback}{%
4265 \edef\hyxmp@today@xmp{\expandafter\hyxmp@pdf@to@xmp@date\pdffeedback creationdate}%
4266 }
4267 % \end{macrocode}
4268 % \begin{macro}{\hyxmp@timestamp}
4269 % Case 3: |\filemoddate| is defined (\XeLaTeX).
4270 % In this case, we treat the timestamp of the job's |.log| file as the
4271 % current date/time.
4272 % \label{page:jobname-log}
4273 % \changes{v5.1}{2020/04/08}{Don't rely on
4274 % \string\cs{jobname}\string\texttt{.aux} existing to query the
4275 % current time under \string\XeLaTeX\string\@. Instead, use
4276 % \string\cs{jobname}\string\texttt{.log}. Thanks to Ulrike Fischer
4277 % for the bug report and for her suggestion to use the log file.}
4278 % \begin{macrocode}
4279 \hyxmp@try@today{filemoddate}{%
4280 \edef\hyxmp@today@xmp{\filemoddate{\[email protected]}}%
4281 \edef\next{%
4282 \edef\noexpand\hyxmp@today@xmp{\noexpand\hyxmp@as@xmp@date{\hyxmp@today@xmp}}%
4283 }%
4284 \next
4285 }%
4286 % \end{macrocode}
4287 % Case 4: None of the above.
4288 % Do the best we can using the available \tex\ primitives (|\year|,
4289 % |\month|, |\day|, and |\time|.
4290 % \begin{macrocode}
4291 \hyxmp@try@today{year}{%
4292 \hyxmp@today@xmp@define\hyxmp@today@xmp
4293 }
4294 % \end{macrocode}
4295 % \end{macro}
4296 % \end{macro}
4297 %
4298 % \begin{macro}{\hyxmp@today@pdf}
4299 % Define |\hyxmp@today@pdf| as the current date and (if available) time
4300 % and timezone in \acro{PDF} date format~\cite{Adobe2008:PDF}. To do so
4301 % we simply convert |\hyxmp@today@xmp|, defined above, from \acro{XMP}
4302 % to \acro{PDF} using |\hyxmp@xmp@to@pdf@date|.
4303 % \changes{v5.0}{2020/03/16}{Added this macro}
4304 % \begin{macrocode}
4305 \expandafter\edef\expandafter\hyxmp@today@pdf\expandafter{%
4306 \expandafter\hyxmp@xmp@to@pdf@date\expandafter{\hyxmp@today@xmp}%
4307 }
4308 % \end{macrocode}
4309 % \end{macro}
4310 %
4311 %
4312 % \subsubsection{Trimming leading and trailing spaces}
4313 % \label{sec:trim-spaces}
4314 %
4315 % To make it easier for \acro{XMP} processors to manipulate our output we
4316 % define a |\hyxmp@trimspaces| macro to strip leading and trailing
4317 % spaces from various data fields.
4318 %
4319 % \begin{macro}{\hyxmp@trimspaces}
4320 % Redefine a macro as its previous value but without leading or trailing
4321 % spaces. This code---as well as that for its helper macros,
4322 % |\hyxmp@trimb| and |\hyxmp@trimc|---was taken almost verbatim from a
4323 % solution to an \emph{Around the Bend} puzzle~\cite{Downes1994:ATB15}.
4324 % Inline comments are also taken from the solution text.
4325 % \begin{macrocode}
4326 \catcode`\Q=3
4327 % \end{macrocode}
4328 % |\hyxmp@trimspaces\x| redefines |\x| to have the same replacement text
4329 % sans leading and trailing space tokens.
4330 % \begin{macrocode}
4331 \newcommand{\hyxmp@trimspaces}[1]{%
4332 % \end{macrocode}
4333 % Use grouping to emulate a multi-token |afterassignment| queue.
4334 % \begin{macrocode}
4335 \begingroup
4336 % \end{macrocode}
4337 % Put ``|\toks 0 {|'' into the |afterassignment| queue.
4338 % \begin{macrocode}
4339 \aftergroup\toks\aftergroup0\aftergroup{%
4340 % \end{macrocode}
4341 % Apply |\hyxmp@trimb| to the replacement text of |#1|, adding a leading
4342 % |\noexpand| to prevent brace stripping and to serve another purpose
4343 % later.
4344 % \begin{macrocode}
4345 \expandafter\hyxmp@trimb\expandafter\noexpand#1Q Q}%
4346 % \end{macrocode}
4347 % Transfer the trimmed text back into |#1|.
4348 % \begin{macrocode}
4349 \edef#1{\the\toks0}%
4350 }
4351 % \end{macrocode}
4352 % \end{macro}
4353 %
4354 % \begin{macro}{\hyxmp@trimb}
4355 % |\hyxmp@trimb| removes a trailing space if present, then calls
4356 % |\hyxmp@trimc| to clean up any leftover bizarre |Q|s, and trim a
4357 % leading space. In order for |\hyxmp@trimc| to work properly we need to
4358 % put back a |Q| first.
4359 % \begin{macrocode}
4360 \def\hyxmp@trimb#1 Q{\hyxmp@trimc#1Q}
4361 % \end{macrocode}
4362 % \end{macro}
4363 %
4364 % \begin{macro}{\hyxmp@trimc}
4365 % Execute |\vfuzz| assignment to remove leading space; the |\noexpand|
4366 % will now prevent unwanted expansion of a macro or other expandable
4367 % token at the beginning of the trimmed text. The |\endgroup| will feed
4368 % in the |\aftergroup| tokens after the |\vfuzz| assignment is
4369 % completed.
4370 % \begin{macrocode}
4371 \def\hyxmp@trimc#1Q#2{\afterassignment\endgroup \vfuzz\the\vfuzz#1}
4372 \catcode`\Q=11
4373 % \end{macrocode}
4374 % \end{macro}
4375 %
4376 %
4377 % \subsubsection{Converting text to XML}
4378 % \label{sec:text-xml}
4379 %
4380 % The ``|<|'', ``|>|'', and ``|&|'' characters are significant to \acro{XML}.
4381 % We therefore need to escape them in any author-supplied text.
4382 %
4383 % \begin{macro}{\ifhyxmp@unicodetex}
4384 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4385 % \begin{macro}{\hyxmp@unicodetextrue}
4386 % \begin{macro}{\hyxmp@unicodetexfalse}
4387 % \XeTeX\ and \LuaTeX\ natively support \term{Unicode}. We define the
4388 % conditional |\ifhyxmp@unicodetex| to check for these so we can
4389 % properly handle encoding conversions. The trick here is that
4390 % \term{Unicode} \tex\ implementations compare decimal~64 to
4391 % hexadecimal~40 (decimal~64), specified with four carets, and take the
4392 % \textsc{true} branch; non-\term{Unicode} \tex\ implementations compare
4393 % decimal~64 to character~``|^|'' (decimal~94), ignore the ``|^^0040|''
4394 % and the rest of the \textsc{true} branch, and take the \textsc{false}
4395 % branch.
4396 % \begin{macrocode}
4397 \newif\ifhyxmp@unicodetex
4398 \ifnum64=`\^^^^0040\relax
4399 \hyxmp@unicodetextrue
4400 \else
4401 \hyxmp@unicodetexfalse
4402 \fi
4403 % \end{macrocode}
4404 % \end{macro}
4405 % \end{macro}
4406 % \end{macro}
4407 %
4408 % \begin{macro}{\SE->pdfdoc@03}
4409 % Preserve \acro{ETX} (|^^C|), which is normally an invalid character in
4410 % \term{PDFDocEncoding}. We use it in \pkgname{hyperxmp} (and
4411 % specifically in |\hyxmp@xmlify| below) as a list-element separator.
4412 % \begin{macrocode}
4413 \expandafter\def\csname SE->pdfdoc@03\endcsname{0003}
4414 % \end{macrocode}
4415 % \end{macro}
4416 %
4417 % \begin{macro}{\SE->pdfdoc@15}
4418 % Preserve \acro{NAK} (|^^U|), which is normally an invalid character in
4419 % \term{PDFDocEncoding}. We use it in \pkgname{hyperxmp} (and
4420 % specifically in |\hyxmp@xmlify| below) as a placeholder for an
4421 % underscore character.
4422 % \begin{macrocode}
4423 \expandafter\def\csname SE->pdfdoc@15\endcsname{0015}
4424 % \end{macrocode}
4425 % \end{macro}
4426 %
4427 % \begin{macro}{\hyxmp@xmlify}
4428 % \changes{v2.0}{2012/08/02}{Completely rewritten by Heiko Oberdiek to
4429 % better support Unicode-enabled \TeX\ programs}
4430 % \begin{macro}{\hyxmp@xmlified}
4431 % \begin{macro}{\hyxmp@text}
4432 % Given a piece of text defined using |\pdfstringdef| (i.e.,~with many
4433 % special characters redefined to have category code~11), set
4434 % |\hyxmp@xmlified| to the same text but with all occurrences of~``|<|''
4435 % replaced with~|&lt;|, all occurrences of~``|>|'' replaced with~|&gt;|,
4436 % and all occurrences of~``|&|'' replaced with~|&amp;|.
4437 % \begin{macrocode}
4438 \newcommand*{\hyxmp@xmlify}[1]{%
4439 \gdef\hyxmp@xmlified{}%
4440 % \end{macrocode}
4441 % Escaped \acro{PDF} string~$\rightarrow$ \term{PDFDocEncoding}/\term{Unicode}
4442 % \begin{macrocode}
4443 \EdefUnescapeString\hyxmp@text{#1}%
4444 \ifhyxmp@unicodetex
4445 % \end{macrocode}
4446 % \term{PDFDocEncoding}/\term{Unicode}~$\rightarrow$ \acro{UTF-32BE}
4447 % \begin{macrocode}
4448 \hyxmp@is@unicode\hyxmp@text{%
4449 \StringEncodingConvert
4450 \hyxmp@text\hyxmp@text{utf16be}{utf32be}%
4451 }{%
4452 \ifXeTeX
4453 \hyxmp@xetex@crap
4454 \else
4455 \StringEncodingConvert
4456 \hyxmp@text\hyxmp@text{pdfdoc}{utf32be}%
4457 \fi
4458 }%
4459 % \end{macrocode}
4460 % \acro{UTF-32BE}~$\rightarrow$ \acro{UTF-32BE} as hex string
4461 % \begin{macrocode}
4462 \EdefEscapeHex\hyxmp@text{\hyxmp@text}%
4463 % \end{macrocode}
4464 % \acro{UTF-32BE}~$\rightarrow$ \acro{XML} in \acro{ASCII}
4465 % \begin{macrocode}
4466 \edef\hyxmp@text{%
4467 \expandafter
4468 }\expandafter\hyxmp@toxml@unicodetex\hyxmp@text
4469 \relax\relax\relax\relax\relax\relax\relax\relax
4470 \else
4471 % \end{macrocode}
4472 % \term{PDFDocEncoding}/\term{Unicode}~$\rightarrow$ \acro{UTF-8}
4473 % \begin{macrocode}
4474 \hyxmp@is@unicode\hyxmp@text{%
4475 \StringEncodingConvert
4476 \hyxmp@text\hyxmp@text{utf16be}{utf8}%
4477 }{%
4478 \StringEncodingConvert
4479 \hyxmp@text\hyxmp@text{pdfdoc}{utf8}%
4480 }%
4481 % \end{macrocode}
4482 % \acro{UTF-8}~$\rightarrow$ \acro{UTF-8} as hex string
4483 % \begin{macrocode}
4484 \EdefEscapeHex\hyxmp@text{\hyxmp@text}%
4485 % \end{macrocode}
4486 % \acro{UTF-8} as hex string~$\rightarrow$ \acro{XML} in \acro{UTF-8}
4487 % as hex string
4488 % \begin{macrocode}
4489 \edef\hyxmp@text{%
4490 \expandafter\hyxmp@toxml\hyxmp@text\@empty\@empty
4491 }%
4492 % \end{macrocode}
4493 % \acro{XML} in \acro{UTF-8} as hex string~$\rightarrow$ \acro{XML}
4494 % in \acro{UTF-8}
4495 % \begin{macrocode}
4496 \EdefUnescapeHex\hyxmp@text{\hyxmp@text}%
4497 \fi
4498 \global\let\hyxmp@xmlified\hyxmp@text
4499 }
4500 % \end{macrocode}
4501 % \end{macro}
4502 % \end{macro}
4503 % \end{macro}
4504 %
4505 % \begin{macro}{\hyxmp@is@unicode}
4506 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4507 % \begin{macro}{\hyxmp@@is@unicode}
4508 % Given a string and two expressions, evaluate the first expression if
4509 % the string is \acro{UTF-16BE}-encoded and the second expression if not.
4510 % \begin{macrocode}
4511 \begingroup
4512 \lccode`\<=254 %
4513 \lccode`\>=255 %
4514 \catcode254=12 %
4515 \catcode255=12 %
4516 \lowercase{\endgroup
4517 \def\hyxmp@is@unicode#1{%
4518 \expandafter\hyxmp@@is@unicode#1<>\@nil
4519 }%
4520 \def\hyxmp@@is@unicode#1<>#2\@nil{%
4521 \ifx\\#1\\%
4522 \expandafter\@firstoftwo
4523 \else
4524 \expandafter\@secondoftwo
4525 \fi
4526 }%
4527 }
4528 % \end{macrocode}
4529 % \end{macro}
4530 % \end{macro}
4531 %
4532 % \begin{macro}{\hyxmp@toxml}
4533 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4534 % Replace the characters ``\textless'', ``\&'', and ``\textgreater'' with
4535 % \acro{XML} entities when using a non-native-\term{Unicode}
4536 % \tex\ (\tex\ or \pdfTeX).
4537 % \begin{macrocode}
4538 \def\hyxmp@toxml#1#2{%
4539 \ifx#1\@empty
4540 \else
4541 \ifnum"#1#2=`\& %
4542 26616D703B% &amp;
4543 \else\ifnum"#1#2=`\< %
4544 266C743B% &lt;
4545 \else\ifnum"#1#2=`\> %
4546 2667743B% &gt;
4547 \else
4548 % \end{macrocode}
4549 % \cmdname{dvips} wraps text when generating most PostScript code but
4550 % preserves line breaks within strings. Unfortunately, \cmdname{dvips}
4551 % fails to observe the special case in the PostScript specification that
4552 % ``[b]alanced pairs of parentheses in the string require no special
4553 % treatment''~\cite{Adobe1996:postscript}. Consequently, \acro{XMP}
4554 % data containing parentheses (e.g.,~``\texttt{Copyright (C) 1605 Miguel
4555 % de Cervantes}'') confuse \cmdname{dvips} into thinking that the
4556 % string has ended after the closing parenthesis and that line breaks
4557 % can subsequently be injected safely into the document at arbitrary
4558 % points for formatting purposes. This leads to erroneous display by
4559 % \acro{PDF} viewers, which honor line breaks within \acro{XMP} tags.
4560 % The solution is to insert a backslash before all parentheses when in
4561 % \texttt{pdfmark}-generating mode to convince \cmdname{dvips} that the
4562 % entire \acro{XMP} packet must be treated as a single,
4563 % not-to-be-modified string.
4564 % \changes{v2.0}{2012/08/22}{Escaped parentheses written with
4565 % \texttt{pdfmark}s to prevent \cmdname{dvips} from line-wrapping the
4566 % \protect\acro{XMP} packet}
4567 % \begin{macrocode}
4568 \@ifundefined{pdfmark}{%
4569 #1#2%
4570 }{%
4571 \ifnum"#1#2=`\( %
4572 5C28% \(
4573 \else\ifnum"#1#2=`\) %
4574 5C29% \)
4575 \else
4576 #1#2%
4577 \fi\fi
4578 }%
4579 \fi\fi\fi
4580 \expandafter\hyxmp@toxml
4581 \fi
4582 }
4583 % \end{macrocode}
4584 % \end{macro}
4585 %
4586 % \begin{macro}{\hyxmp@toxml@unicodetex}
4587 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4588 % \begin{macro}{\hyxmp@text}
4589 % Replace the characters ``\textless'', ``\&'', and ``\textgreater'' with
4590 % \acro{XML} entities when using a native-\term{Unicode}
4591 % \tex\ (\XeTeX\ or \LuaTeX).
4592 % \begin{macrocode}
4593 \def\hyxmp@toxml@unicodetex#1#2#3#4#5#6#7#8{%
4594 \ifx#1\relax
4595 \else
4596 \ifnum"#1#2#3#4#5#6#7#8>127 %
4597 \uccode`\*="#1#2#3#4#5#6#7#8\relax
4598 \uppercase{%
4599 \edef\hyxmp@text{\hyxmp@text *}%
4600 }%
4601 \else\ifnum"#7#8=`\< %
4602 \edef\hyxmp@text{\hyxmp@text &lt;}%
4603 \else\ifnum"#7#8=`\& %
4604 \edef\hyxmp@text{\hyxmp@text &amp;}%
4605 \else\ifnum"#7#8=`\> %
4606 \edef\hyxmp@text{\hyxmp@text &gt;}%
4607 \else\ifnum"#7#8=`\ %
4608 \edef\hyxmp@text{\hyxmp@text\space}%
4609 \else
4610 \uccode`\*="#7#8\relax
4611 \uppercase{%
4612 \edef\hyxmp@text{\hyxmp@text *}%
4613 }%
4614 \fi\fi\fi\fi\fi
4615 \expandafter\hyxmp@toxml@unicodetex
4616 \fi
4617 }
4618 % \end{macrocode}
4619 % \end{macro}
4620 % \end{macro}
4621 %
4622 % \begin{macro}{\hyxmp@skipzeros}
4623 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4624 % Skip over leading zeroes in the input argument.
4625 % \begin{macrocode}
4626 \def\hyxmp@skipzeros#1{%
4627 \ifx#10%
4628 \expandafter\hyxmp@skipzeros
4629 \fi
4630 }
4631 % \end{macrocode}
4632 % \end{macro}
4633 %
4634 % \begin{macro}{\x}
4635 % \begin{macro}{\hyxmp@xetex@crap}
4636 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4637 % \begin{macro}{\hyxmp@try}
4638 % \begin{macro}{\hyxmp@crap@result}
4639 % \begin{macro}{\hyxmp@text}
4640 % In the case of \XeTeX, the strings defined by |\pdfstringdef| can
4641 % contain big characters. In this case, the string is treated as
4642 % \term{Unicode}.
4643 % \begin{macrocode}
4644 \begingroup
4645 \def\x#1{\endgroup
4646 \def\hyxmp@xetex@crap{%
4647 \edef\hyxmp@try{%
4648 \expandafter\hyxmp@SpaceOther\hyxmp@text#1\@nil
4649 }%
4650 \let\hyxmp@crap@result=N%
4651 \expandafter\hyxmp@crap@test\hyxmp@try\relax
4652 \ifx\hyxmp@crap@result Y%
4653 \let\hyxmp@text\@empty
4654 \expandafter\hyxmp@crap@convert\hyxmp@try\relax
4655 \else
4656 \StringEncodingConvert\hyxmp@text\hyxmp@text{pdfdoc}{utf32be}%
4657 \fi
4658 }%
4659 }
4660 \x{ }
4661 % \end{macrocode}
4662 % \end{macro}
4663 % \end{macro}
4664 % \end{macro}
4665 % \end{macro}
4666 % \end{macro}
4667 %
4668 % \begin{macro}{\hyxmp@SpaceOther}
4669 % Re-encode all spaces in a string with category code~12 (``other'').
4670 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4671 % \begin{macrocode}
4672 \begingroup
4673 \catcode`\~=12 %
4674 \lccode`\~=`\ %
4675 \lowercase{\endgroup
4676 \def\hyxmp@SpaceOther#1 #2\@nil{%
4677 #1%
4678 \ifx\relax#2\relax
4679 \expandafter\@gobble
4680 \else
4681 ~%
4682 \expandafter\@firstofone
4683 \fi
4684 {\hyxmp@SpaceOther#2\@nil}%
4685 }%
4686 }
4687 % \end{macrocode}
4688 % \end{macro}
4689 %
4690 %
4691 % \begin{macro}{\hyxmp@crap@test}
4692 % Determine if we need to treat a string as \term{Unicode}.
4693 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4694 % \begin{macrocode}
4695 \def\hyxmp@crap@test#1{%
4696 \ifx#1\relax
4697 \else
4698 \ifnum`#1>127 %
4699 \let\hyxmp@crap@result=Y%
4700 \expandafter\expandafter\expandafter\hyxmp@skiptorelax
4701 \else
4702 \expandafter\expandafter\expandafter\hyxmp@crap@test
4703 \fi
4704 \fi
4705 }
4706 % \end{macrocode}
4707 % \end{macro}
4708 %
4709 % \begin{macro}{\hyxmp@skiptorelax}
4710 % Discard all tokens up to and including the first |\relax|.
4711 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4712 % \begin{macrocode}
4713 \def\hyxmp@skiptorelax#1\relax{}
4714 % \end{macrocode}
4715 % \end{macro}
4716 %
4717 % \begin{macro}{\hyxmp@crap@convert}
4718 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4719 % \begin{macro}{\hyxmp@num}
4720 % \begin{macro}{\hyxmp@text}
4721 % Convert a hexadecimal string to a number.
4722 % \begin{macrocode}
4723 \def\hyxmp@crap@convert#1{%
4724 \ifx#1\relax
4725 \else
4726 \edef\hyxmp@num{\number`#1}%
4727 \ifnum\hyxmp@num>"FFFFFF %
4728 \lccode`\!=\intcalcDiv{\hyxmp@num}{\number"1000000}\relax
4729 \lowercase{\edef\hyxmp@text{\hyxmp@text!}}%
4730 \edef\hyxmp@num{\intcalcMod{\hyxmp@num}{\number"1000000}}%
4731 \else
4732 \edef\hyxmp@text{\hyxmp@text\hyxmp@zero}%
4733 \fi
4734 \ifnum\hyxmp@num>"FFFF %
4735 \lccode`\!=\intcalcDiv{\hyxmp@num}{\number"10000}\relax
4736 \lowercase{\edef\hyxmp@text{\hyxmp@text!}}%
4737 \edef\hyxmp@num{\intcalcMod{\hyxmp@num}{\number"10000}}%
4738 \else
4739 \edef\hyxmp@text{\hyxmp@text\hyxmp@zero}%
4740 \fi
4741 \ifnum\hyxmp@num>"FF %
4742 \lccode`\!=\intcalcDiv{\hyxmp@num}{\number"100}\relax
4743 \lowercase{\edef\hyxmp@text{\hyxmp@text!}}%
4744 \edef\hyxmp@num{\intcalcMod{\hyxmp@num}{\number"100}}%
4745 \else
4746 \edef\hyxmp@text{\hyxmp@text\hyxmp@zero}%
4747 \fi
4748 \ifnum\hyxmp@num>0 %
4749 \lccode`\!=\hyxmp@num\relax
4750 \lowercase{\edef\hyxmp@text{\hyxmp@text!}}%
4751 \else
4752 \edef\hyxmp@text{\hyxmp@text\hyxmp@zero}%
4753 \fi
4754 \expandafter\hyxmp@crap@convert
4755 \fi
4756 }
4757 % \end{macrocode}
4758 % \end{macro}
4759 % \end{macro}
4760 % \end{macro}
4761 %
4762 % \begin{macro}{\hyxmp@zero}
4763 % Define a null character with category code~12 (``other'').
4764 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
4765 % \begin{macrocode}
4766 \begingroup
4767 \catcode0=12 %
4768 \gdef\hyxmp@zero{^^00}%
4769 \endgroup
4770 % \end{macrocode}
4771 % \end{macro}
4772 %
4773 %
4774 % \subsubsection{Outputting structured XML}
4775 % \label{sec:output-xml}
4776 %
4777 % An \acro{XMP} packet consists of structured \acro{XML} data. We
4778 % define some helper routines to handle the repetitive tasks of
4779 % indenting a consistent number of spaces, inserting begin and end tags,
4780 % and escaping arbitrary text as necessary for \acro{XML} compatibility.
4781 %
4782 % \begin{macro}{\hyxmp@extra@indent}
4783 % This macro is used internally to increase the amount of indentation
4784 % when writing certain \acro{XML} data. It is normally defined as empty
4785 % but can temporarily be redefined to a sequence of |\space| characters.
4786 % \begin{macrocode}
4787 \newcommand*{\hyxmp@extra@indent}{}
4788 % \end{macrocode}
4789 % \end{macro}
4790 %
4791 % \begin{macro}{\hyxmp@add@simple}
4792 % Given an \acro{XMP} tag~(|#1|) and a string~(|#2|), if the string is
4793 % nonempty, add a begin tag, the string, and an end tag to the packet.
4794 % The ``|simple|'' in the macro name indicates that the string is output
4795 % without variations for different languages.
4796 % \changes{v2.0}{2012/08/25}{Added this macro}
4797 % \changes{v5.0}{2020/02/21}{Insert the tag name (\string\texttt{\#1})
4798 % verbatim}
4799 % \begin{macrocode}
4800 \newcommand*{\hyxmp@add@simple}[2]{%
4801 \@ifnotmtargexp{#2}{%
4802 \hyxmp@xmlify{#2}%
4803 \hyxmp@add@to@xml{\hyxmp@extra@indent______<}%
4804 \xdef\hyxmp@xml{\hyxmp@xml#1}%
4805 \hyxmp@add@to@xml{>\hyxmp@xmlified</}%
4806 \xdef\hyxmp@xml{\hyxmp@xml#1>^^J}%
4807 }%
4808 }
4809 % \end{macrocode}
4810 % \end{macro}
4811 %
4812 % \begin{macro}{\hyxmp@add@simple@var}
4813 % Given an \acro{XMP} tag~(|#1|) and a variable name~(|#2|), if the
4814 % string is defined, add a begin tag, the string, and an end tag to the
4815 % packet. The ``|simple|'' in the macro name indicates that the string
4816 % is output without variations for different languages.
4817 % |\hyxmp@add@simple@var| differs from |\hyxmp@add@simple| in that the
4818 % former includes defined but empty values in the \acro{XMP} packet
4819 % while the latter excludes both undefined and defined but empty values.
4820 % \changes{v2.4}{2013/12/21}{Added this macro}
4821 % \begin{macrocode}
4822 \newcommand*{\hyxmp@add@simple@var}[2]{%
4823 \expandafter\ifx\csname#2\endcsname\relax
4824 \else
4825 \hyxmp@xmlify{\csname#2\endcsname}%
4826 \hyxmp@add@to@xml{%
4827 \hyxmp@extra@indent______<#1>\hyxmp@xmlified</#1>^^J%
4828 }%
4829 \fi
4830 }
4831 % \end{macrocode}
4832 % \end{macro}
4833 %
4834 % \begin{macro}{\hyxmp@add@simple@lang}
4835 % Given an \acro{XMP} tag~(|#1|) and a string~(|#2|), if the string is
4836 % nonempty, add a begin tag, the string, and an end tag to the packet.
4837 % The ``|simple|'' in the macro name indicates that the string is output
4838 % without variations for different languages. However, if the string
4839 % begins with a language code in square brackets, specify that as the
4840 % (sole) language for the tag.
4841 % \changes{v4.0}{2019/03/12}{Added this macro}
4842 % \begin{macrocode}
4843 \newcommand*{\hyxmp@add@simple@lang}[2]{%
4844 \@ifnotmtarg{#2}{%
4845 \hyxmp@xmlify{#2}%
4846 \expandafter\hyxmp@add@simple@lang@i\hyxmp@xmlified\relax{#1}%
4847 }%
4848 }
4849 % \end{macrocode}
4850 % \end{macro}
4851 %
4852 % \begin{macro}{\hyxmp@add@simple@lang@i}
4853 % This is a helper macro for |\hyxmp@add@simple@lang|. It takes an
4854 % optional language code (in brackets), text up to |\relax|, and a tag,
4855 % and typesets the text within the \acro{XML} tag.
4856 % \begin{macrocode}
4857 \newcommand*{\hyxmp@add@simple@lang@i}{%
4858 \@ifnextchar[\hyxmp@add@simple@lang@ii{\hyxmp@add@simple@lang@ii[\@pdfmetalang]}%
4859 }
4860 % \end{macrocode}
4861 % \end{macro}
4862 %
4863 % \begin{macro}{\hyxmp@add@simple@lang@ii}
4864 % This is another helper macro for |\hyxmp@add@simple@lang|. It takes
4865 % an mandatory language code (in brackets; can be empty), text up to
4866 % |\relax|, and a tag, and typesets the text within the \acro{XML} tag.
4867 % \begin{macrocode}
4868 \def\hyxmp@add@simple@lang@ii[#1]#2\relax#3{%
4869 \@ifnotmtarg{#2}{%
4870 \hyxmp@xmlify{#2}%
4871 \@ifmtarg{#1}{%
4872 \hyxmp@add@to@xml{%
4873 ______<#3>\hyxmp@xmlified</#3>^^J%
4874 }%
4875 }{%
4876 \hyxmp@add@to@xml{%
4877 ______<#3 xml:lang="#1">\hyxmp@xmlified</#3>^^J%
4878 }%
4879 }%
4880 }%
4881 }
4882 % \end{macrocode}
4883 % \end{macro}
4884 %
4885 % \begin{macro}{\hyxmp@add@simple@pfx}
4886 % Given an \acro{XMP} tag~(|#1|), a---typically hard-wired---prefix
4887 % string~(|#2|), and a main string~(|#2|), if the main string is
4888 % nonempty, add a begin tag, both strings, and an end tag to the packet.
4889 % The ``|simple|'' in the macro name indicates that the string is output
4890 % without variations for different languages.
4891 % \changes{v5.2}{2020/04/29}{Added this macro}
4892 % \begin{macrocode}
4893 \newcommand*{\hyxmp@add@simple@pfx}[3]{%
4894 \@ifnotmtargexp{#3}{%
4895 \hyxmp@add@to@xml{\hyxmp@extra@indent______<}%
4896 \xdef\hyxmp@xml{\hyxmp@xml#1}%
4897 \hyxmp@pdfstringdef\hyxmp@iprefix{#2}%
4898 \hyxmp@xmlify{\hyxmp@iprefix}%
4899 \hyxmp@add@to@xml{>\hyxmp@xmlified}%
4900 \hyxmp@xmlify{#3}%
4901 \hyxmp@add@to@xml{\hyxmp@xmlified</}%
4902 \xdef\hyxmp@xml{\hyxmp@xml#1>^^J}%
4903 }%
4904 }
4905 % \end{macrocode}
4906 % \end{macro}
4907 %
4908 %
4909 % \subsubsection{Providing metadata in multiple languages}
4910 % \label{sec:lang-alt}
4911 %
4912 % Certain \acro{XMP} tags---\xmpprop{dc:title},
4913 % \xmpprop{dc:description}, and \xmpprop{dc:rights} (and others? Let me
4914 % know.)---can be expressed in multiple languages. The same text is
4915 % used for both language \optname{pdfmetalang} (default:
4916 % \optname{pdflang}) and language ``\xmpterm{x-default}''. To express
4917 % the same metadata in multiple languages, we provide an |\XMPLangAlt|
4918 % macro to construct a list of alternative forms for a piece of
4919 % metadata.
4920 %
4921 % \begin{macro}{\hyxmp@alt@title}
4922 % \begin{macro}{\hyxmp@alt@description}
4923 % \begin{macro}{\hyxmp@alt@rights}
4924 % Each of these macros is a list in which each element is of the form
4925 % ``|\do| \meta{language} \meta{text}'' in which \meta{language} is an
4926 % \acro{ISO} \mbox{639-1} two-letter country code with an optional
4927 % \acro{ISO} \mbox{3166-1} two-letter region code. For example,
4928 % |\hyxmp@alt@title| may contain an element, ``|\do| |{es-MX}|
4929 % |{Este| |es| |mi| |documento}|''.
4930 % \begin{macrocode}
4931 \def\hyxmp@alt@title{}
4932 \def\hyxmp@alt@description{}
4933 \def\hyxmp@alt@rights{}
4934 % \end{macrocode}
4935 % \end{macro}
4936 % \end{macro}
4937 % \end{macro}
4938 %
4939 % \begin{macro}{\hyxmp@LA@accept}
4940 % This macro wraps |\define@key| to make the option ``|#1|=\meta{value}''
4941 % append \meta{value} to list~|#2|.
4942 % \begin{macrocode}
4943 \newcommand{\hyxmp@LA@accept}[2]{%
4944 \define@key{hyxmp@LA}{#1}{%
4945 % \end{macrocode}
4946 % \begin{macro}{\hyxmp@value}
4947 % As Niklas Beisert observed, if the option passed to the current key
4948 % contains \LaTeX\ code, this code will be included in the \acro{XMP}
4949 % packet, which is undesirable. Hence, we first clean up the string
4950 % using |\hyxmp@pdfstringdef|.
4951 % \begin{macrocode}
4952 \hyxmp@pdfstringdef\hyxmp@value{##1}%
4953 \xdef#2{#2\noexpand\do {\hyxmp@cur@lang} {\hyxmp@value}}%
4954 }
4955 }
4956 % \end{macrocode}
4957 % \end{macro}
4958 % \end{macro}
4959 %
4960 % Define \meta{key}=\meta{value} options for appending to each of the
4961 % |\hyxmp@alt|\meta{tag} lists.
4962 % \begin{macrocode}
4963 \hyxmp@LA@accept{pdftitle}{\hyxmp@alt@title}
4964 \hyxmp@LA@accept{pdfsubject}{\hyxmp@alt@description}
4965 \hyxmp@LA@accept{pdfcopyright}{\hyxmp@alt@rights}
4966 % \end{macrocode}
4967 %
4968 % \begin{macro}{\XMPLangAlt}
4969 % Argument~|#1| is a language expressed as a two-letter country code and
4970 % optional two-letter region code. Argument~|#2| is a list of
4971 % \meta{key}=\meta{value} pairs. Keys correspond to |\hypersetup|
4972 % options such as ``|pdftitle|'', ``|pdfsubject|'', and
4973 % ``|pdfcopyright|''. Values are the alternative-language form of the
4974 % text provided for the corresponding option.
4975 % \changes{v3.3}{2017/07/22}{Added this macro based on a request---and
4976 % some code---by Niklas Beisert to support metadata expressed in multiple
4977 % languages}
4978 % \begin{macrocode}
4979 \newcommand{\XMPLangAlt}[2]{%
4980 \let\do=\relax
4981 % \end{macrocode}
4982 % \begin{macro}{\hyxmp@cur@lang}
4983 % Store the provided language, which will be used during option processing.
4984 % \begin{macrocode}
4985 \edef\hyxmp@cur@lang{#1}%
4986 \setkeys{hyxmp@LA}{#2}%
4987 }
4988 % \end{macrocode}
4989 % \end{macro}
4990 % \end{macro}
4991 %
4992 %
4993 % \subsection{UUID generation}
4994 % \label{sec:uuid-gen}
4995 %
4996 % We use a linear congruential generator to produce pseudorandom
4997 % version~4 \acro{UUID}s~\cite{Leach2005:uuid}. True, this method has
4998 % its flaws but it's simple to implement in \tex\ and is good enough for
4999 % producing the \acro{XMP} \xmpprop{xmpMM:DocumentID} and
5000 % \xmpprop{xmpMM:InstanceID} fields.
5001 %
5002 % \begin{macro}{\hyxmp@modulo@a}
5003 % Replace the contents of |\@hyxmp@count| with the contents modulo~|#1|.
5004 % Note that |\@tempcntb| is overwritten in the process.
5005 % \begin{macrocode}
5006 \def\hyxmp@modulo@a#1{%
5007 \@tempcntb=\@hyxmp@count
5008 \divide\@tempcntb by #1
5009 \multiply\@tempcntb by #1
5010 \advance\@hyxmp@count by -\@tempcntb
5011 }
5012 % \end{macrocode}
5013 % \end{macro}
5014 %
5015 % \begin{macro}{\hyxmp@big@prime}
5016 % \begin{macro}{\hyxmp@big@prime@ii}
5017 % Define a couple of large prime numbers that can still be stored in a
5018 % \tex\ counter.
5019 % \begin{macrocode}
5020 \def\hyxmp@big@prime{536870923}
5021 \def\hyxmp@big@prime@ii{536870027}
5022 % \end{macrocode}
5023 % \end{macro}
5024 % \end{macro}
5025 %
5026 % \begin{macro}{\hyxmp@seed@rng}
5027 % \begin{macro}{\hyxmp@one@token}
5028 % Seed \pkgname{hyperxmp}'s random-number generator from a given piece
5029 % of text.
5030 % \begin{macrocode}
5031 \def\hyxmp@seed@rng#1{%
5032 \@hyxmp@count=\hyxmp@big@prime
5033 \futurelet\hyxmp@one@token\hyxmp@seed@rng@i#1\@empty
5034 }
5035 % \end{macrocode}
5036 % \end{macro}
5037 % \end{macro}
5038 %
5039 % \begin{macro}{\hyxmp@seed@rng@i}
5040 % \begin{macro}{\hyxmp@one@token}
5041 % \begin{macro}{\next}
5042 % Do all of the work for |\hyxmp@seed@rng|. For each character code $c$
5043 % of the input text, assign $\mathtt{\string\@hyxmp@count} \leftarrow 3
5044 % \cdot \mathtt{\string\@hyxmp@count} + c
5045 % \pmod{\mathtt{\string\hyxmp@big@prime}}$.
5046 % \begin{macrocode}
5047 \def\hyxmp@seed@rng@i{%
5048 \ifx\hyxmp@one@token\@empty
5049 \let\next=\relax
5050 \else
5051 \def\next##1{%
5052 \multiply\@hyxmp@count by 3
5053 \advance\@hyxmp@count by `##1
5054 \hyxmp@modulo@a{\hyxmp@big@prime}%
5055 \futurelet\hyxmp@one@token\hyxmp@seed@rng@i
5056 }%
5057 \fi
5058 \next
5059 }
5060 % \end{macrocode}
5061 % \end{macro}
5062 % \end{macro}
5063 % \end{macro}
5064 %
5065 % \begin{macro}{\hyxmp@set@rand@num}
5066 % \begin{macro}{\hyxmp@rand@num}
5067 % Advance |\hyxmp@rand@num| to the next pseudorandom number in the
5068 % sequence. Specifically, we assign $\mathtt{\string\hyxmp@rand@num}
5069 % \leftarrow 3 \cdot \mathtt{\string\hyxmp@rand@num} +
5070 % \mathtt{\string\hyxmp@big@prime@ii}
5071 % \pmod{\mathtt{\string\hyxmp@big@prime}}$. Note that both |\@hyxmp@count|
5072 % and |\@tempcntb| are overwritten in the process.
5073 % \begin{macrocode}
5074 \def\hyxmp@set@rand@num{%
5075 \@hyxmp@count=\hyxmp@rand@num
5076 \multiply\@hyxmp@count by 3
5077 \advance\@hyxmp@count by \hyxmp@big@prime@ii
5078 \hyxmp@modulo@a{\hyxmp@big@prime}%
5079 \xdef\hyxmp@rand@num{\the\@hyxmp@count}%
5080 }
5081 % \end{macrocode}
5082 % \end{macro}
5083 % \end{macro}
5084 %
5085 % \begin{macro}{\hyxmp@append@hex}
5086 % Append a randomly selected hexadecimal digit to macro~|#1|. Note that
5087 % both |\@hyxmp@count| and |\@tempcntb| are overwritten in the process.
5088 % \begin{macrocode}
5089 \def\hyxmp@append@hex#1{%
5090 \hyxmp@set@rand@num
5091 \@hyxmp@count=\hyxmp@rand@num
5092 \hyxmp@modulo@a{16}%
5093 \ifnum\@hyxmp@count<10
5094 \xdef#1{#1\the\@hyxmp@count}%
5095 \else
5096 % \end{macrocode}
5097 % There \emph{must} be a better way to handle the numbers~10--15 than
5098 % with |\ifcase|.
5099 % \begin{macrocode}
5100 \advance\@hyxmp@count by -10
5101 \ifcase\@hyxmp@count
5102 \xdef#1{#1a}%
5103 \or\xdef#1{#1b}%
5104 \or\xdef#1{#1c}%
5105 \or\xdef#1{#1d}%
5106 \or\xdef#1{#1e}%
5107 \or\xdef#1{#1f}%
5108 \fi
5109 \fi
5110 }
5111 % \end{macrocode}
5112 % \end{macro}
5113 %
5114 % \begin{macro}{\hyxmp@append@hex@iii}
5115 % Invoke |\hyxmp@append@hex| three times.
5116 % \begin{macrocode}
5117 \def\hyxmp@append@hex@iii#1{%
5118 \hyxmp@append@hex#1%
5119 \hyxmp@append@hex#1%
5120 \hyxmp@append@hex#1%
5121 }
5122 % \end{macrocode}
5123 % \end{macro}
5124 %
5125 % \begin{macro}{\hyxmp@append@hex@iv}
5126 % Invoke |\hyxmp@append@hex| four times.
5127 % \begin{macrocode}
5128 \def\hyxmp@append@hex@iv#1{%
5129 \hyxmp@append@hex@iii#1%
5130 \hyxmp@append@hex#1%
5131 }
5132 % \end{macrocode}
5133 % \end{macro}
5134 %
5135 % \begin{macro}{\hyxmp@create@uuid}
5136 % As per the definition of a version~4
5137 % \acro{UUID}~\cite{Leach2005:uuid}, define macro~|#1| as a \acro{UUID}
5138 % of the form
5139 % ``|uuid:|\textit{xxxxxxxx}|-|\linebreak[0]\textit{xxxx}|-|\linebreak[0]\textit{4xxx}|-|\linebreak[0]\textit{yxxx}|-|\linebreak[0]\textit{xxxxxxxxxxxx}''
5140 % in which each ``\textit{x}'' is a lowercase hexadecimal digit and
5141 % ``\textit{y}'' is one of ``|8|'', ``|9|'', ``|a|'', or~``|b|''. We
5142 % assume that the random-number generator is already seeded. Note that
5143 % |\hyxmp@create@uuid| overwrites both |\@hyxmp@count| and |\@tempcntb|.
5144 % \changes{v2.4}{2014/01/02}{Modified this macro to produce a proper
5145 % version~4 (random or pseudorandom) \protect\acro{UUID}}
5146 % \begin{macrocode}
5147 \def\hyxmp@create@uuid#1{%
5148 \def#1{uuid:}%
5149 \hyxmp@append@hex@iv#1%
5150 \hyxmp@append@hex@iv#1%
5151 \g@addto@macro#1{-}%
5152 \hyxmp@append@hex@iv#1%
5153 \g@addto@macro#1{-4}%
5154 \hyxmp@append@hex@iii#1%
5155 \g@addto@macro#1{-}%
5156 % \end{macrocode}
5157 % Randomly select one of ``|8|'', ``|9|'', ``|a|'', or~``|b|''.
5158 % \begin{macrocode}
5159 \hyxmp@set@rand@num
5160 \@hyxmp@count=\hyxmp@rand@num
5161 \hyxmp@modulo@a{4}%
5162 \ifcase\@hyxmp@count
5163 \g@addto@macro#1{8}%
5164 \or\g@addto@macro#1{9}%
5165 \or\g@addto@macro#1{a}%
5166 \or\g@addto@macro#1{b}%
5167 \fi
5168 \hyxmp@append@hex@iii#1%
5169 \g@addto@macro#1{-}%
5170 \hyxmp@append@hex@iv#1%
5171 \hyxmp@append@hex@iv#1%
5172 \hyxmp@append@hex@iv#1%
5173 }
5174 % \end{macrocode}
5175 % \end{macro}
5176 %
5177 % \begin{macro}{\hyxmp@def@DocumentID}
5178 % \begin{macro}{\hyxmp@DocumentID}
5179 % \begin{macro}{\hyxmp@seed@string}
5180 % Seed the random-number generator with a function of the current
5181 % filename, \acro{PDF} document title, and \acro{PDF} author, then invoke
5182 % |\hyxmp@create@uuid| to define |\hyxmp@DocumentID| as a random \acro{UUID}.
5183 % \changes{v3.4}{2017/11/04}{Correctly handle an author field of all spaces.
5184 % Bug reported by Ga\"etan Leurent}
5185 % \begin{macrocode}
5186 \newcommand*{\hyxmp@def@DocumentID}{%
5187 \edef\hyxmp@seed@string{\hyxmp@jobname:\@pdftitle:\@pdfauthor:}%
5188 \expandafter\hyxmp@seed@rng\expandafter{\hyxmp@seed@string}%
5189 \edef\hyxmp@rand@num{\the\@hyxmp@count}%
5190 \hyxmp@create@uuid\hyxmp@DocumentID
5191 }
5192 % \end{macrocode}
5193 % \end{macro}
5194 % \end{macro}
5195 % \end{macro}
5196 %
5197 % \begin{macro}{\hyxmp@def@InstanceID}
5198 % \begin{macro}{\hyxmp@InstanceID}
5199 % \begin{macro}{\hyxmp@seed@string}
5200 % Seed the random-number generator with a function of the current
5201 % filename, \acro{PDF} document title, \acro{PDF} author, and the
5202 % current timestamp, then invoke |\hyxmp@create@uuid| to define
5203 % |\hyxmp@InstanceID| as a random \acro{UUID}. For the current
5204 % timestamp, we use both the document-specified timestamp from
5205 % \optname{pdfdate} and the \tex\ time. The former can be more precise
5206 % (to sub-seconds) but may be less random (as it depends on manual
5207 % document modifications) while the latter is typically less precise (to
5208 % minutes) but may be more random (as it is updated automatically).
5209 % \changes{v3.5}{2018/11/27}{Seed with the \protect\TeX\ timestamp in
5210 % addition to the document-specified timestamp}
5211 % \begin{macrocode}
5212 \newcommand*{\hyxmp@def@InstanceID}{%
5213 \hyxmp@today@xmp@define{\hyxmp@seed@string}%
5214 \edef\hyxmp@seed@string{%
5215 \hyxmp@jobname:\@pdftitle:\@pdfauthor:\hyxmp@today@xmp:\hyxmp@seed@string
5216 }%
5217 \expandafter\hyxmp@seed@rng\expandafter{\hyxmp@seed@string}%
5218 \edef\hyxmp@rand@num{\the\@hyxmp@count}%
5219 \hyxmp@create@uuid\hyxmp@InstanceID
5220 }
5221 % \end{macrocode}
5222 % \end{macro}
5223 % \end{macro}
5224 % \end{macro}
5225 %
5226 %
5227 % \subsection{Constructing the XMP packet}
5228 %
5229 % An \acro{XMP} packet ``shall consist of the following, in order: a
5230 % header \acro{PI}, the serialized \acro{XMP} data model (the \acro{XMP}
5231 % packet) with optional white-space padding, and a trailer
5232 % \acro{PI}''~\cite{Adobe2012:XMP}. (``\acro{PI}'' is an abbreviation for
5233 % ``processing instructions''). The serialized \acro{XMP} includes blocks
5234 % of \acro{XML} for various \acro{XMP} schemata:
5235 % ^^A
5236 % \schema[Adobe PDF]{Adobe \protect\acro{PDF}} (Section~\ref{sec:adobe-pdf}),
5237 % \schema{Dublin Core} (Section~\ref{sec:dublin-core}),
5238 % \schema[XMP Rights Management]{\protect\acro{XMP} Rights Management} (Section~\ref{sec:xmp-rights}),
5239 % \schema[XMP Media Management]{\protect\acro{XMP} Media Management} (Section~\ref{sec:xmp-media}),
5240 % \schema[XMP Basic]{\protect\acro{XMP} Basic} (Section~\ref{sec:xmp-basic}),
5241 % \schema{Photoshop} (Section~\ref{sec:photoshop}),
5242 % \index{PDF/A Identification schema=\acro {PDF/A} Identification schema}
5243 % \index{PDF/X Identification schema=\acro {PDF/X} Identification schema}
5244 % \index{PDF/UA Identification schema=\acro {PDF/UA} Identification schema}
5245 % \index{schemata&PDF/A Identification=\acro {PDF/A} Identification}
5246 % \index{schemata&PDF/X Identification=\acro {PDF/X} Identification}
5247 % \index{schemata&PDF/UA Identification=\acro {PDF/UA} Identification}
5248 % \acro{PDF}\textsc{/*} Identification (Section~\ref{sec:pdfX-id}),
5249 % \schema[IPTC Photo Metadata]{\protect\acro{IPTC} Photo Metadata} (Section~\ref{sec:photo-meta}),
5250 % \schema[PRISM Basic Metadata]{\protect\acro{PRISM} Basic Metadata} (Section~\ref{sec:prism-meta}),
5251 % \schema{Journal Article Versions} (Section~\ref{sec:niso-jav-meta}), and
5252 % \schema[XMP Paged-Text]{\protect\acro{XMP} Paged-Text} (Section~\ref{sec:xmp-paged-text}).
5253 % ^^A
5254 % The |\hyxmp@construct@packet| macro
5255 % (Section~\ref{sec:combining-schemata}) constructs the \acro{XMP} packet
5256 % into |\hyxmp@xml|. It first writes the appropriate \acro{XML} header,
5257 % then calls the various schema-writing macros, then injects
5258 % |\hyxmp@padding| as padding, and finally writes the appropriate
5259 % \acro{XML} trailer.
5260 %
5261 % \subsubsection{XMP utility functions}
5262 %
5263 % \begin{macro}{\hyxmp@add@to@xml}
5264 % Given a piece of text, replace all underscores with category-code~11
5265 % (``other'') spaces and all |^C| characters with commas, then append
5266 % the result to the |\hyxmp@xml| macro.
5267 % \changes{v2.0}{2012/08/24}{Updated also to replace commas}
5268 % \changes{v2.5}{2014/06/19}{Updated also to replace underscores and to
5269 % modify only the text being added, not the already-modified text}
5270 % \changes{v2.8}{2014/04/05}{Corrected inadvertent lowercasing of non-Latin
5271 % characters when run under \XeLaTeX\ or \LuaLaTeX\ (bug reported by
5272 % Leonid Sinev)}
5273 % \begin{macrocode}
5274 \newcommand*{\hyxmp@add@to@xml}[1]{%
5275 \bgroup
5276 \@hyxmp@count=0
5277 \ifhyxmp@unicodetex
5278 \@tempcntb=65536%
5279 \else
5280 \@tempcntb=256%
5281 \fi
5282 \loop
5283 \lccode\@hyxmp@count=\@hyxmp@count
5284 \advance\@hyxmp@count by 1
5285 \ifnum\@hyxmp@count<\@tempcntb
5286 \repeat
5287 \lccode`\_=`\ \relax
5288 \lccode`\^^C=`\,\relax
5289 \lccode`\^^U=`\_\relax
5290 \lowercase{\xdef\hyxmp@new@xml{#1}}%
5291 \xdef\hyxmp@xml{\hyxmp@xml\hyxmp@new@xml}%
5292 \egroup
5293 }
5294 % \end{macrocode}
5295 % \end{macro}
5296 %
5297 % \begin{macro}{\hyxmp@hash}
5298 % Define a category-code~11 (``other'') version of the ``|#|'' character.
5299 % \begin{macrocode}
5300 \bgroup
5301 \catcode`\#=11
5302 \gdef\hyxmp@hash{#}
5303 \egroup
5304 % \end{macrocode}
5305 % \end{macro}
5306 %
5307 % \begin{macro}{\hyxmp@padding}
5308 % \begin{macro}{\hyxmp@xml}
5309 % The \acro{XMP} specification recommends leaving approximately
5310 % 2000~bytes of whitespace at the end of each \acro{XMP} packet to
5311 % facilitate editing the packet in place~\cite{Adobe2012:XMP}.
5312 % |\hyxmp@padding| is defined to contain 32~lines of 63~spaces and a
5313 % newline apiece for a total of 2048 characters of whitespace.
5314 % \begin{macrocode}
5315 \bgroup
5316 \xdef\hyxmp@xml{}%
5317 \hyxmp@add@to@xml{%
5318 _______________________________________________________________^^J%
5319 }
5320 \xdef\hyxmp@padding{\hyxmp@xml}%
5321 \egroup
5322 \xdef\hyxmp@padding{\hyxmp@padding\hyxmp@padding}
5323 \xdef\hyxmp@padding{\hyxmp@padding\hyxmp@padding}
5324 \xdef\hyxmp@padding{\hyxmp@padding\hyxmp@padding}
5325 \xdef\hyxmp@padding{\hyxmp@padding\hyxmp@padding}
5326 \xdef\hyxmp@padding{\hyxmp@padding\hyxmp@padding}
5327 % \end{macrocode}
5328 % \end{macro}
5329 % \end{macro}
5330 %
5331 % \begin{macro}{\hyxmp@x@default}
5332 % Define an \xmpterm{x-default} string that we can use in comparisons
5333 % with |\@pdfmetalang|.
5334 % \begin{macrocode}
5335 \newcommand*{\hyxmp@x@default}{x-default}
5336 % \end{macrocode}
5337 % \end{macro}
5338 %
5339 %
5340 % \subsubsection{The Adobe PDF schema}
5341 % \label{sec:adobe-pdf}
5342 %
5343 % \begin{schemadesc}[Adobe PDF]{Adobe \protect\acro{PDF}}
5344 %
5345 % Older versions of \pkgname{hyperref} defined a default producer; newer
5346 % versions do not. Instead, they let the \tex\ engine define the
5347 % producer itself. This poses a problem for \acro{PDF/A} compliance
5348 % because \pkgname{hyperxmp} sees an empty producer and therefore omits
5349 % writing a \xmpprop{pdf:Producer} to the \acro{XMP} packet, causing a
5350 % mismatch between the data in the \acro{XMP} packet and the data in the
5351 % \acro{PDF} \pdfterm{Info} dictionary. To ensure consistency between
5352 % \acro{XMP} and \pdfterm{Info}, we explicitly define our own default
5353 % |\@pdfproducer| here.
5354 %
5355 % \begin{macro}{\@pdfproducer}
5356 % \begin{macro}{\hyxmp@define@pdfproducer}
5357 % Define |\@pdfproducer| using the banner string if available or the
5358 % \tex\ engine's version number if not.
5359 % \changes{v5.0}{2020/03/01}{Added this macro}
5360 % \changes{v5.1}{2020/04/06}{Check for \string\LuaTeX\ before checking for
5361 % \string\pdfTeX\ to work around \string\pkgname{luatex85}'s confusing
5362 % \string\pkgname{iftex} by defining \string\cs{pdftexversion}. Thanks
5363 % to Robin Schwab for the bug report}
5364 % \begin{macrocode}
5365 \newcommand*{\hyxmp@define@pdfproducer}{%
5366 \gdef\@pdfproducer{TeX}
5367 \ifLuaTeX
5368 \expandafter\hyxmp@banner@to@producer\expandafter{\luatexbanner}%
5369 \else
5370 \ifPDFTeX
5371 \expandafter\hyxmp@banner@to@producer\expandafter{\pdftexbanner}%
5372 \else
5373 \ifXeTeX
5374 \edef\@pdfproducer{XeTeX version \the\XeTeXversion\XeTeXrevision}%
5375 \fi
5376 \fi
5377 \fi
5378 }
5379 % \end{macrocode}
5380 % \end{macro}
5381 % \end{macro}
5382 %
5383 % \begin{macro}{\@pdfproducer}
5384 % \begin{macro}{\hyxmp@banner@to@producer}
5385 % Define |\@pdfproducer| as the \tex\ engine's banner string
5386 % \ifPDFTeX
5387 % (e.g.,~``\texttt{\pdftexbanner}''),
5388 % \else
5389 % \ifLuaTeX
5390 % (e.g.,~``\texttt{\luatexbanner}''),
5391 % \else
5392 % (e.g.,~``\texttt{This is pdfTeX, Version 3.14159265-2.6-1.40.20
5393 % (TeX Live 2019/Debian) kpathsea version 6.3.1}''),
5394 % \fi
5395 % \fi
5396 % removing the initial ``\texttt{This is}'' if possible (specifically,
5397 % when $\varepsilon$-\TeX's\index{e-TeX=$\varepsilon$-\TeX}
5398 % |\scantokens| primitive is available).
5399 % \changes{v5.1}{2020/04/03}{Prevent the category code of ``@'' from
5400 % propagating past the \cs{begin}\string\texttt{\{document\}}.
5401 % Thanks to Robert Schlicht for noticing this catcode ``leak'' and
5402 % providing a correction}
5403 % \begin{macrocode}
5404 \def\hyxmp@banner@to@producer#1{%
5405 \ifx\scantokens\relax
5406 \gdef\@pdfproducer{#1}%
5407 \else
5408 {\scantokens{\makeatletter\hyxmp@remove@this#1\relax}}%
5409 \fi
5410 }
5411 % \end{macrocode}
5412 % \end{macro}
5413 % \end{macro}
5414 %
5415 % \begin{macro}{\@pdfproducer}
5416 % \begin{macro}{\hyxmp@remove@this}
5417 % Define |\@pdfproducer| as a given banner string with the initial
5418 % ``\texttt{This is}'' stripped off the beginning.
5419 % \begin{macrocode}
5420 \def\hyxmp@remove@this This is #1\relax{\gdef\@pdfproducer{#1}}
5421 % \end{macrocode}
5422 % \end{macro}
5423 % \end{macro}
5424 %
5425 % If \optname{pdfproducer} wasn't specified and \pkgname{hyperref}
5426 % didn't already define |\@pdfproducer|---old versions of
5427 % \pkgname{hyperref} did; newer ones don't---try to assign a meaningful
5428 % producer string and use that.
5429 % \changes{v5.0}{2020/03/02}{Define a default producer}
5430 % \begin{macrocode}
5431 \AtBeginDocument{%
5432 \ifx\@pdfproducer\relax
5433 \hyxmp@define@pdfproducer
5434 \fi
5435 }
5436 % \end{macrocode}
5437 %
5438 % \begin{macro}{\hyxmp@assign@major@minor}
5439 % Assign |\hyxmp@major@minor| to be the \acro{PDF} version targeted by
5440 % the running \TeX\ engine.
5441 % \changes{v5.2}{2020/05/10}{Added this macro. \string\pkgname{hyperxmp}
5442 % now correctly specifies \string\xmpprop{pdf:PDFVersion} when
5443 % generating \string\acro{PDF}~2.0+. Thanks to Ulrike Fischer for
5444 % alerting me to \string\acro{PDF}~2.0's availability in the
5445 % \string\TeX\ ecosystem and informing me how to activate it}
5446 % \begin{macro}{\hyxmp@major@minor}
5447 % \begin{macrocode}
5448 \newcommand*{\hyxmp@assign@major@minor}{%
5449 \@ifundefined{pdfvariable}{%
5450 \@ifundefined{pdfminorversion}{%
5451 % \end{macrocode}
5452 % Case 1: Neither |\pdfvariable| nor |\pdfminorversion| is defined
5453 % (\XeLaTeX\ and regular \LaTeX).
5454 % \begin{macrocode}
5455 }{%
5456 % \end{macrocode}
5457 % Case 2: |\pdfminorversion| is defined (\pdfLaTeX\ and pre-0.85 \LuaLaTeX).
5458 % \begin{macrocode}
5459 \xdef\hyxmp@major@minor{\the\pdfminorversion}%
5460 \@ifundefined{pdfmajorversion}{%
5461 % \end{macrocode}
5462 % Case 2(a): |\pdfmajorversion| is not defined (older versions of
5463 % \pdfLaTeX\ and \LuaLaTeX).
5464 % \begin{macrocode}
5465 \xdef\hyxmp@major@minor{1.\hyxmp@major@minor}%
5466 }{%
5467 % \end{macrocode}
5468 % Case 2(b): |\pdfmajorversion| is defined (\pdfLaTeX~1.40.21+).
5469 % \begin{macrocode}
5470 \xdef\hyxmp@major@minor{\the\pdfmajorversion.\hyxmp@major@minor}%
5471 }%
5472 }%
5473 }{%
5474 % \end{macrocode}
5475 % Case 3: |\pdfvariable| is defined (\LuaLaTeX~0.85+).
5476 % \begin{macrocode}
5477 \xdef\hyxmp@major@minor{\the\pdfvariable majorversion.\the\pdfvariable minorversion}%
5478 }%
5479 }
5480 % \end{macrocode}
5481 % \end{macro}
5482 % \end{macro}
5483 %
5484 % \begin{macro}{\hyxmp@pdf@schema}
5485 % Add properties defined by the Adobe \acro{PDF} schema to the |\hyxmp@xml|
5486 % macro.
5487 % \begin{macrocode}
5488 \newcommand*{\hyxmp@pdf@schema}{%
5489 % \end{macrocode}
5490 % Add a block of \acro{XML} to |\hyxmp@xml| that lists the document's
5491 % keywords (the \xmpprop{pdf:Keywords} property), the tools used to
5492 % produce the \acro{PDF} file (the \xmpprop{pdf:Producer} property), and
5493 % the version of the \acro{PDF} standard adhered to (the
5494 % \xmpprop{pdf:PDFVersion} property). Unlike most of the other schemata
5495 % that \pkgname{hyperxmp} supports, the Adobe \acro{PDF} schema is
5496 % \emph{always} included in the document, even if all of its keys are
5497 % empty. This is because \PDFstd{A}{1}{b}{} requires the keywords and
5498 % producer to be the same in the \acro{XMP} metadata and the \acro{PDF}
5499 % metadata. Because \pkgname{hyperref} always specifies the
5500 % \pdfterm{Keywords} and \pdfterm{Producer} fields, even when they're
5501 % empty, \pkgname{hyperxmp} has to follow suit and define
5502 % \xmpprop{pdf:Keywords} and \xmpprop{pdf:Producer} in the \acro{XMP}
5503 % packet.
5504 % \changes{v2.4}{2013/12/21}{Made
5505 % \texttt{\string\string\string\hyxmp@pdf@schema}
5506 % \protect\emph{always} include the Adobe \protect\acro{PDF} schema,
5507 % even when empty. Florian Breitwieser noted that this is necessary
5508 % for \protect\PDFstd{A}{1}{b}{} compliance}
5509 % \changes{v5.0}{2020/02/28}{Honor \protect\optname{pdftrapped}}
5510 % \begin{macrocode}
5511 \hyxmp@add@simple@var{pdf:Producer}{@pdfproducer}%
5512 \hyxmp@add@simple@var{pdf:Keywords}{@pdfkeywords}%
5513 \hyxmp@add@simple{pdf:Trapped}{\@pdftrapped}%
5514 \hyxmp@assign@major@minor
5515 \hyxmp@add@simple@var{pdf:PDFVersion}{hyxmp@major@minor}%
5516 }
5517 % \end{macrocode}
5518 % \end{macro}
5519 %
5520 % \end{schemadesc}
5521 %
5522 %
5523 % \subsubsection{The Dublin Core schema}
5524 % \label{sec:dublin-core}
5525 %
5526 % \begin{schemadesc}{Dublin Core}
5527 %
5528 % \begin{macro}{\ifhyxmp@multi@langs}
5529 % \begin{macro}{\hyxmp@multi@langstrue}
5530 % \begin{macro}{\hyxmp@multi@langsfalse}
5531 % These macros are used locally to \cs{hyxmp@rdf@dc}. If the property
5532 % being processed has values in different languages,
5533 % \cs{ifhyxmp@multi@langs} evaluates to \textsc{true}. If it has a
5534 % value in only a single language, \cs{ifhyxmp@multi@langs} evaluates to
5535 % \textsc{false}.
5536 % \begin{macrocode}
5537 \newif\ifhyxmp@multi@langs
5538 % \end{macrocode}
5539 % \end{macro}
5540 % \end{macro}
5541 % \end{macro}
5542 %
5543 % \begin{macro}{\hyxmp@rdf@dc}
5544 % Given an optional |\if|\meta{something} statement (|#1|), a Dublin Core
5545 % property~(|#2|) and a macro containing some |\pdfstringdef|-defined
5546 % text~(|#3|), append the appropriate block of \acro{XML} to the |\hyxmp@xml|
5547 % macro.
5548 % \changes{v1.4}{2011/06/12}{Included metadata in the \xmpterm{x-default}
5549 % language regardless of the specified metadata language}
5550 % \changes{v3.3}{2017/07/22}{Bug fix: Output the metadata language as
5551 % correct \noexpand\acro{XML} even when \noexpand\pkgname{hyperref} is
5552 % loaded with the \noexpand\optname{unicode} option}
5553 % \changes{v5.5}{2020/07/17}{List \protect\xmpterm{x-default} alternatives
5554 % before language-specific alternatives, as dictated by the
5555 % \protect\acro{XMP} specification~\protect\cite{Adobe2012:XMP}}
5556 % \changes{v5.5}{2020/07/21}{Rewrite the core part of this macro to
5557 % divide it into four, cleanly defined cases}
5558 % \begin{macrocode}
5559 \newcommand*{\hyxmp@rdf@dc}[3][\iffalse]{%
5560 % \end{macrocode}
5561 % Set |\@tempswatrue| only if the given text is nonempty or the provided
5562 % conditional evaluates to \textsc{true}.
5563 % \begin{macrocode}
5564 \@ifmtargexp{#3}{\@tempswafalse}{\@tempswatrue}%
5565 #1
5566 \@tempswatrue
5567 \fi
5568 % \end{macrocode}
5569 % Append the corresponding \acro{XML} only if |\@tempswatrue|.
5570 % \begin{macrocode}
5571 \if@tempswa
5572 \hyxmp@xmlify{#3}%
5573 % \end{macrocode}
5574 % \begin{macro}{\hyxmp@value}
5575 % Store the \acro{XML}-ified version of |#3| in |\hyxmp@value| so we can
5576 % reuse |\hyxmp@xmlifiied| if necessary.
5577 % \begin{macrocode}
5578 \let\hyxmp@value=\hyxmp@xmlified
5579 \hyxmp@add@to@xml{%
5580 ______<dc:#2>^^J%
5581 ________<rdf:Alt>^^J%
5582 }%
5583 % \end{macrocode}
5584 % Record whether property~|#2| has definitions in multiple languages.
5585 % \begin{macrocode}
5586 \@if@def@and@nonempty{hyxmp@alt@#2}{%
5587 \hyxmp@multi@langstrue
5588 }{%
5589 \hyxmp@multi@langsfalse
5590 }%
5591 % \end{macrocode}
5592 % There are now four cases to consider: the cross product of
5593 % \{\optname{pdfmetalang}~$=$ ``\xmpterm{x-default}'',
5594 % \optname{pdfmetalang}~$\neq$ ``\xmpterm{x-default}''\} and
5595 % \{\cs{XMPLangAlt} was specified, \cs{XMPLangAlt} was not specified\}.
5596 % We handle each of these in turn.
5597 % \begin{macrocode}
5598 \ifx\@pdfmetalang\hyxmp@x@default
5599 \ifhyxmp@multi@langs
5600 % \end{macrocode}
5601 % \textbf{Case~1}: No \optname{pdfmetalang} but \cs{XMPLangAlt}.
5602 % We consider this an error because the \xmpterm{x-default} language will
5603 % not have a matching non-default language, in violation of the
5604 % \acro{XMP} specification's guidance~\cite[p.~23]{Adobe2012:XMP}:
5605 %
5606 % \begin{quote}
5607 % \sffamily
5608 %
5609 % An \textbf{xml:lang} value of ``x-default'' may be used to
5610 % explicitly denote a default item. If used, the ``x-default'' item
5611 % shall be first in the array and its simple text value should be
5612 % repeated in another item in which \textbf{xml:lang} specifies its
5613 % actual language. However, an ``x-default'' item may be the only
5614 % item, in which case there is only a default value in no defined
5615 % language.
5616 % \end{quote}
5617 % \begin{macrocode}
5618 \PackageError{hyperxmp}%
5619 {\string\XMPLangAlt\space was used without specifying
5620 pdfmetalang\MessageBreak
5621 or pdflang}%
5622 {Be sure to assign a language code to the pdfmetalang key and/or a
5623 document\MessageBreak
5624 language to the pdflang key (e.g., \string\hypersetup{pdfmetalang={en}}).}%
5625 \else
5626 % \end{macrocode}
5627 % \textbf{Case~2}: No \optname{pdfmetalang} and no \cs{XMPLangAlt}.
5628 % Here we specify only \xmpterm{x-default} as the language, as per the
5629 % guidance quoted above.
5630 % \begin{macrocode}
5631 \hyxmp@add@to@xml{%
5632 __________<rdf:li xml:lang="\hyxmp@x@default">\hyxmp@value</rdf:li>^^J%
5633 }%
5634 \fi
5635 \else
5636 \ifhyxmp@multi@langs
5637 % \end{macrocode}
5638 % \textbf{Case~3}: Both \optname{pdfmetalang} and \cs{XMPLangAlt}. In
5639 % this case, we include an \xmpterm{x-default} followed by the
5640 % \optname{pdfmetalang} language, followed by all of the language
5641 % alternatives.
5642 % \begin{macrocode}
5643 \hyxmp@xmlify{\@pdfmetalang}%
5644 \hyxmp@add@to@xml{%
5645 __________<rdf:li xml:lang="\hyxmp@x@default">\hyxmp@value</rdf:li>^^J%
5646 __________<rdf:li xml:lang="\hyxmp@xmlified">\hyxmp@value</rdf:li>^^J%
5647 }%
5648 \def\do##1##2{
5649 \hyxmp@xmlify{##2}%
5650 \hyxmp@add@to@xml{%
5651 __________<rdf:li xml:lang="##1">\hyxmp@xmlified</rdf:li>^^J%
5652 }%
5653 }%
5654 \csname hyxmp@alt@#2\endcsname
5655 \else
5656 % \end{macrocode}
5657 % \textbf{Case~4}: \optname{pdfmetalang} but no \cs{XMPLangAlt}. To
5658 % reduce redundancy we omit the \xmpterm{x-default} and include the
5659 % single language in which the text appears.
5660 % \begin{macrocode}
5661 \hyxmp@xmlify{\@pdfmetalang}%
5662 \hyxmp@add@to@xml{%
5663 __________<rdf:li xml:lang="\hyxmp@xmlified">\hyxmp@value</rdf:li>^^J%
5664 }%
5665 \fi
5666 \fi
5667 % \end{macrocode}
5668 % Complete this \acro{XMP} element.
5669 % \begin{macrocode}
5670 \hyxmp@add@to@xml{%
5671 ________</rdf:Alt>^^J%
5672 ______</dc:#2>^^J%
5673 }%
5674 \fi
5675 }%
5676 % \end{macrocode}
5677 % \end{macro}
5678 % \end{macro}
5679 %
5680 % \begin{macro}{\hyxmp@list@to@xml}
5681 % Given an optional |\if|\meta{something} statement (|#1|), a Dublin Core
5682 % property~(|#2|), an \acro{RDF} array~(|#3|), and a macro containing a
5683 % comma-separated list~(|#4|), append the appropriate block of \acro{XML} to
5684 % the |\hyxmp@xml| macro.
5685 % \changes{v2.0}{2012/08/02}{Modified by Heiko Oberdiek to use the new
5686 % \term{Unicode}-processing macros}
5687 % \begin{macrocode}
5688 \newcommand*{\hyxmp@list@to@xml}[4][\iffalse]{%
5689 % \end{macrocode}
5690 % Set |\@tempswatrue| only if the given list is nonempty or the provided
5691 % conditional evaluates to \textsc{true}.
5692 % \begin{macrocode}
5693 \@ifmtargexp{#4}{\@tempswafalse}{\@tempswatrue}%
5694 #1
5695 \@tempswatrue
5696 \fi
5697 % \end{macrocode}
5698 % Append the corresponding \acro{XML} only if |\@tempswatrue|.
5699 % \begin{macrocode}
5700 \if@tempswa
5701 \hyxmp@add@to@xml{%
5702 ______<dc:#2>^^J%
5703 ________<rdf:#3>^^J%
5704 }%
5705 \bgroup
5706 % \end{macrocode}
5707 % \begin{macro}{\@elt}
5708 % Re-encode the text from \term{Unicode} if necessary. Then redefine
5709 % |\@elt| to \acro{XML}-ify each element of the list and append it to
5710 % |\hyxmp@xmlified|.
5711 % \begin{macrocode}
5712 \hyxmp@xmlify{#4}%
5713 \hyxmp@commas@to@list\hyxmp@list{\hyxmp@xmlified}%
5714 \def\@elt##1{%
5715 \hyxmp@add@to@xml{%
5716 __________<rdf:li>##1</rdf:li>^^J%
5717 }%
5718 }%
5719 \hyxmp@list
5720 \egroup
5721 \hyxmp@add@to@xml{%
5722 ________</rdf:#3>^^J%
5723 ______</dc:#2>^^J%
5724 }%
5725 \fi
5726 }
5727 % \end{macrocode}
5728 % \end{macro}
5729 % \end{macro}
5730 %
5731 % \begin{macro}{\hyxmp@singleton@dc}
5732 % Given an optional list type (|Seq| or |Bag|), a Dublin Core property,
5733 % and a string, append a block of \acro{XML} representing a one-element
5734 % list consisting of the given string.
5735 % \changes{v4.1}{2019/04/05}{Added this macro}
5736 % \begin{macrocode}
5737 \newcommand{\hyxmp@singleton@dc}[3][Bag]{%
5738 \@ifnotmtargexp{#3}{%
5739 \hyxmp@xmlify{#3}%
5740 \hyxmp@add@to@xml{%
5741 ______<dc:#2>^^J%
5742 ________<rdf:#1>^^J%
5743 __________<rdf:li>\hyxmp@xmlified</rdf:li>^^J%
5744 ________</rdf:#1>^^J%
5745 ______</dc:#2>^^J%
5746 }%
5747 }
5748 }
5749 % \end{macrocode}
5750 % \end{macro}
5751 %
5752 % \begin{macro}{\hyxmp@cond@dc@identifier}
5753 % Conditionally add a \xmpprop{dc:identifier} tag. Given a prefix
5754 % string~(|#1|) and a main string~(|#2|), wrap these in a
5755 % \xmpprop{dc:identifier} if the main string is nonempty and
5756 % |\hyxmp@xmlified| \emph{is} empty (implying the
5757 % \xmpprop{dc:identifier} has not yet been written).
5758 % \changes{v5.2}{2020/04/29}{Added this macro}
5759 % \begin{macrocode}
5760 \newcommand*{\hyxmp@cond@dc@identifier}[2]{%
5761 \ifx\hyxmp@xmlified\@empty
5762 \@ifnotmtargexp{#2}{%
5763 \hyxmp@add@simple@pfx{dc:identifier}{#1}{#2}%
5764 }%
5765 \fi
5766 }
5767 % \end{macrocode}
5768 % \end{macro}
5769 %
5770 % \begin{macro}{\hyxmp@dc@schema}
5771 % Add properties defined by the Dublin Core schema to the |\hyxmp@xml|
5772 % macro. Specifically, we add entries for the \xmpprop{dc:title}
5773 % property if the author specified a \optname{pdftitle}, the
5774 % \xmpprop{dc:description} property if the author specified a
5775 % \optname{pdfsubject}, the \xmpprop{dc:rights} property if the author
5776 % specified a \optname{pdfcopyright}, the \xmpprop{dc:creator} property
5777 % if the author specified a \optname{pdfauthor}, the
5778 % \xmpprop{dc:subject} property if the author specified
5779 % \optname{pdfkeywords}, the \xmpprop{dc:language} property if the
5780 % author specified \optname{pdflang}, the \xmpprop{dc:type} property if
5781 % the author specified \optname{pdftype}, and the
5782 % \xmpprop{dc:identifier} if the author specified
5783 % \optname{pdfidentifier} or if we can derive it from other options. We
5784 % also specify the \xmpprop{dc:source} property using the base name of
5785 % the source file with |.tex| appended and the \xmpprop{dc:date}
5786 % property using the date the document was run through \LaTeX---unless
5787 % the author specified \optname{pdfdate}, in which case we use that.
5788 % \changes{v2.0}{2012/08/26}{Added support for \xmpprop{dc:language}
5789 % and \xmpprop{dc:source}}
5790 % \changes{v2.4}{2013/12/21}{Made \xmpprop{dc:language} a \xmpterm{Bag}
5791 % instead of an individual item so as to conform to the latest
5792 % \acro{XMP} specifications, a detail identified by Florian Breitwieser}
5793 % \changes{v5.3}{2020/06/13}{Include all languages used in the document
5794 % in \protect\xmpprop{dc:language}}
5795 % \changes{v5.4}{2020/06/18}{Bug fix: Use \protect\cs{hyxmp@today@xmp}
5796 % as the date only if \protect\cs{@pdfdatetime} is undefined}
5797 % \begin{macrocode}
5798 \newcommand*{\hyxmp@dc@schema}{%
5799 \hyxmp@add@simple{dc:format}{application/pdf}%
5800 \hyxmp@rdf@dc[\ifHy@pdfa]{title}{\@pdftitle}%
5801 \hyxmp@rdf@dc[\ifHy@pdfa]{description}{\@pdfsubject}%
5802 \hyxmp@rdf@dc{rights}{\@pdfcopyright}%
5803 \hyxmp@singleton@dc{publisher}{\@pdfpublisher}%
5804 \@ifmtargexp{\@pdfdatetime}{%
5805 \hyxmp@singleton@dc[Seq]{date}{\hyxmp@today@xmp}%
5806 }{%
5807 \hyxmp@singleton@dc[Seq]{date}{\@pdfdatetime}%
5808 }%
5809 \hyxmp@singleton@dc{type}{\@pdftype}%
5810 \hyxmp@list@to@xml[\ifHy@pdfa]{creator}{Seq}{\hyxmp@pdfauthor}%
5811 \hyxmp@list@to@xml{subject}{Bag}{\hyxmp@pdfkeywords}%
5812 \ifx\@pdfsource\@empty
5813 \else
5814 \hyxmp@add@simple{dc:source}{\@pdfsource}%
5815 \fi
5816 \hyxmp@list@to@xml{language}{Bag}{\hyxmp@dc@lang}%
5817 % If |\@pdfidentifier| is empty, try setting it to each of |\@pdfdoi|,
5818 % |\@pdfeissn|, |\@pdfissn|, and |\@pdfisbn|, in turn, with proper
5819 % syntactic adjustments.
5820 % \begin{macrocode}
5821 \@ifmtargexp{\@pdfidentifier}{%
5822 \let\hyxmp@xmlified=\@empty
5823 \hyxmp@cond@dc@identifier{info:doi/}{\@pdfdoi}%
5824 \hyxmp@cond@dc@identifier{urn:ISSN:}{\@pdfeissn}%
5825 \hyxmp@cond@dc@identifier{urn:ISSN:}{\@pdfissn}%
5826 \hyxmp@cond@dc@identifier{urn:ISBN:}{\@pdfisbn}%
5827 }{%
5828 \hyxmp@add@simple{dc:identifier}{\@pdfidentifier}%
5829 }%
5830 }
5831 % \end{macrocode}
5832 % \end{macro}
5833 %
5834 % \end{schemadesc}
5835 %
5836 %
5837 % \subsubsection{The XMP Rights Management schema}
5838 % \label{sec:xmp-rights}
5839 %
5840 % \begin{schemadesc}[XMP Rights Management]{\protect\acro{XMP} Rights Management}
5841 %
5842 % \begin{macro}{\hyxmp@xmpRights@schema}
5843 % Add properties defined by the \acro{XMP} Rights Management schema to the
5844 % |\hyxmp@xml| macro. Currently, these are only the
5845 % \xmpprop{xmpRights:Marked} property and the
5846 % \xmpprop{xmpRights:WebStatement} property. If the author specified a
5847 % copyright statement we mark the document as copyrighted. If the
5848 % author specified a license statement we include the \acro{URL} in the
5849 % metadata.
5850 % \changes{v1.4}{2011/05/29}{Renamed the \texttt{xapRights} namespace
5851 % prefix to \texttt{xmpRights}}
5852 % \changes{v2.0}{2012/08/25}{Modified to include
5853 % \xmpprop{xmpRights:Marked} only when \optname{pdfcopyright} is
5854 % specified and \xmpprop{xmpRights:WebStatement} only when
5855 % \optname{pdflicenseurl} is specified}
5856 % \begin{macrocode}
5857 \newcommand*{\hyxmp@xmpRights@schema}{%
5858 % \end{macrocode}
5859 % \begin{macro}{\hyxmp@legal}
5860 % Set |\hyxmp@rights| to |YES| if either \optname{pdfcopyright} or
5861 % \optname{pdflicenseurl} was specified.
5862 % \begin{macrocode}
5863 \let\hyxmp@rights=\@empty
5864 \ifx\@pdflicenseurl\@empty
5865 \else
5866 \def\hyxmp@rights{YES}%
5867 \fi
5868 \ifx\@pdfcopyright\@empty
5869 \else
5870 \def\hyxmp@rights{YES}%
5871 \fi
5872 % \end{macrocode}
5873 % Include the license-statement \acro{URL} and/or the copyright
5874 % indication. The copyright statement itself is included by
5875 % |\hyxmp@dc@schema| in Section~\ref{sec:dublin-core}.
5876 % \begin{macrocode}
5877 \ifx\hyxmp@rights\@empty
5878 \else
5879 \ifx\@pdfcopyright\@empty
5880 \else
5881 \hyxmp@add@simple{xmpRights:Marked}{True}%
5882 \fi
5883 \hyxmp@add@simple{xmpRights:WebStatement}{\@pdflicenseurl}%
5884 \fi
5885 }
5886 % \end{macrocode}
5887 % \end{macro}
5888 % \end{macro}
5889 %
5890 % \end{schemadesc}
5891 %
5892 %
5893 % \subsubsection{The XMP Media Management schema}
5894 % \label{sec:xmp-media}
5895 %
5896 % \begin{schemadesc}[XMP Media Management]{\protect\acro{XMP} Media Management}
5897 %
5898 % \begin{macro}{\hyxmp@aep@toks}
5899 % Once we reach the end of the preamble and know that \cs{@pdftitle} and
5900 % \cs{@pdfauthor} are no longer expected to change we use those macros
5901 % (and others) to define one \acro{UUID} for the document
5902 % (\cs{hyxmp@DocumentID}) and one for the document instance
5903 % (\cs{hyxmp@InstanceID}). As explained in
5904 % Section~\ref{sec:initial-prep}, we defer the invocation of
5905 % \cs{AtEndPreamble} to the end of the file.
5906 % \changes{v5.7}{2020/10/13}{As requested by Moritz Heckscher, define
5907 % \protect\cs{hyxmp@DocumentID} and \protect\cs{hyxmp@InstanceID} at
5908 % the end of the preamble instead of at the end of the document}
5909 % \begin{macrocode}
5910 \expandafter\hyxmp@aep@toks\expandafter=\expandafter{%
5911 \the\hyxmp@aep@toks
5912 \AtEndPreamble{%
5913 \@ifmtargexp{\hyxmp@DocumentID}{\hyxmp@def@DocumentID}{}%
5914 \@ifmtargexp{\hyxmp@InstanceID}{\hyxmp@def@InstanceID}{}%
5915 }%
5916 }
5917 % \end{macrocode}
5918 % \end{macro}
5919 %
5920 % \begin{macro}{\hyxmp@mm@schema}
5921 % Add properties defined by the \acro{XMP} Media Management schema to
5922 % the |\hyxmp@xml| macro. According to the \acro{XMP} specification,
5923 % the \xmpprop{xmpMM:DocumentID} property is supposed to uniquely
5924 % identify a document, and the \xmpprop{xmpMM:InstanceID} property is
5925 % supposed to change with each save operation~\cite{Adobe2012:XMP}. As
5926 % seen in Section~\ref{sec:uuid-gen}, we do what we can to honor this
5927 % intention from within a \tex-based workflow. We additionally support
5928 % the \xmpprop{xmpMM:VersionID} property, whose value is supplied by
5929 % the author using \optname{pdfversionid}.
5930 % \changes{v1.4}{2011/05/29}{Renamed the \texttt{xapMM} namespace
5931 % prefix to \texttt{xmpMM}}
5932 % \changes{v3.5}{2018/11/27}{Generate \protect\cs{hyxmp@DocumentID} and
5933 % \protect\cs{hyxmp@InstanceID} only if the document does not already
5934 % define these using the \protect\optname{pdfdocumentid} and
5935 % \protect\optname{pdfinstanceid} options}
5936 % \changes{v4.0}{2019/03/09}{Include \protect\xmpprop{xmpMM:VersionID} in
5937 % the \protect\acro{XMP} packet}
5938 % \begin{macrocode}
5939 \gdef\hyxmp@mm@schema{%
5940 \hyxmp@add@simple{xmpMM:DocumentID}{\hyxmp@DocumentID}%
5941 \hyxmp@add@simple{xmpMM:InstanceID}{\hyxmp@InstanceID}%
5942 \hyxmp@add@simple{xmpMM:VersionID}{\@pdfversionid}%
5943 \hyxmp@add@simple{xmpMM:RenditionClass}{\@pdfrendition}%
5944 }
5945 % \end{macrocode}
5946 % \end{macro}
5947 %
5948 % \end{schemadesc}
5949 %
5950 %
5951 % \subsubsection{The XMP Basic schema}
5952 % \label{sec:xmp-basic}
5953 %
5954 % \begin{schemadesc}[XMP Basic]{\protect\acro{XMP} Basic}
5955
5956 % \begin{macro}{\hyxmp@xmp@basic@schema}
5957 % Add properties defined by the \acro{XMP} Basic schema to the
5958 % |\hyxmp@xml| macro. These include a bunch of dates (all set to the
5959 % same value) and the base \acro{URL} for the document if specified with
5960 % \optname{baseurl}.
5961 % \changes{v2.0}{2012/08/26}{Added this macro}
5962 % \changes{v3.0}{2016/07/04}{Made the \protect\acro{XMP}
5963 % \protect\xmpprop{xmp:CreateDate}, \protect\xmpprop{xmp:ModifyDate},
5964 % and \protect\xmpprop{xmp:MetadataDate} match the \protect\acro{PDF}
5965 % \protect\pdfterm{CreationDate}}
5966 % \changes{v3.2}{2017/01/22}{Honor \string\pkgname{hyperref}'s
5967 % \string\optname{pdfcreationdate} and \string\optname{pdfmoddate}
5968 % options plus a new \string\optname{pdfmetadate} option. Leonid
5969 % Sinev requested this additional control and helped test the resulting
5970 % \protect\pkgname{hyperxmp} code}
5971 % \begin{macrocode}
5972 \newcommand*{\hyxmp@xmp@basic@schema}{%
5973 % \end{macrocode}
5974 % For the document's creation date, use the user-specified
5975 % |\@pdfcreationdate| if defined and non-empty. Otherwise use our
5976 % fabricated |\hyxmp@today@xmp|.
5977 % \begin{macrocode}
5978 \@ifmtargexp{\@pdfcreationdate}{%
5979 \hyxmp@add@simple{xmp:CreateDate}{\hyxmp@today@xmp}%
5980 }{%
5981 \hyxmp@add@simple{xmp:CreateDate}{%
5982 \expandafter\hyxmp@as@xmp@date\expandafter{\@pdfcreationdate}}%
5983 }%
5984 % \end{macrocode}
5985 % For the document's modification date, use the user-specified
5986 % |\@pdfmoddate| if defined and non-empty. Otherwise use our
5987 % fabricated |\hyxmp@today@xmp|.
5988 % \begin{macrocode}
5989 \@ifmtargexp{\@pdfmoddate}{%
5990 \hyxmp@add@simple{xmp:ModifyDate}{\hyxmp@today@xmp}%
5991 }{%
5992 \hyxmp@add@simple{xmp:ModifyDate}{%
5993 \expandafter\hyxmp@as@xmp@date\expandafter{\@pdfmoddate}}%
5994 }%
5995 % \end{macrocode}
5996 % For the document's metadata date, use the user-specified
5997 % |\@pdfmetadatetime| if defined and non-empty. Otherwise use our
5998 % fabricated |\hyxmp@today@xmp|.
5999 % \begin{macrocode}
6000 \@ifmtargexp{\@pdfmetadatetime}{%
6001 \hyxmp@add@simple{xmp:MetadataDate}{\hyxmp@today@xmp}%
6002 }{%
6003 \hyxmp@add@simple{xmp:MetadataDate}{\@pdfmetadatetime}%
6004 }%
6005 % \end{macrocode}
6006 % Define the creation tool and the base \acro{URL}.
6007 % \begin{macrocode}
6008 \hyxmp@add@simple{xmp:CreatorTool}{\@pdfcreator}%
6009 \hyxmp@add@simple{xmp:BaseURL}{\@baseurl}%
6010 }
6011 % \end{macrocode}
6012 % \end{macro}
6013 %
6014 % \end{schemadesc}
6015 %
6016 %
6017 % \subsubsection{The Photoshop schema}
6018 % \label{sec:photoshop}
6019 %
6020 % \begin{schemadesc}{Photoshop}
6021 %
6022 % \begin{macro}{\hyxmp@photoshop@schema}
6023 % \changes{v2.0}{2012/08/26}{Simplified using
6024 % \texttt{\string\string\string\hyxmp@add@simple}}
6025 % \begin{macro}{\hyxmp@photoshop@data}
6026 % Add properties defined by the Photoshop schema to the |\hyxmp@xml|
6027 % macro. We currently support only the
6028 % \xmpprop{photoshop:AuthorsPosition} and
6029 % \xmpprop{photoshop:CaptionWriter} properties.
6030 % \begin{macrocode}
6031 \gdef\hyxmp@photoshop@schema{%
6032 \edef\hyxmp@photoshop@data{\@pdfauthortitle\@pdfcaptionwriter}%
6033 \hyxmp@add@simple{photoshop:AuthorsPosition}{\@pdfauthortitle}%
6034 \hyxmp@add@simple{photoshop:CaptionWriter}{\@pdfcaptionwriter}%
6035 }
6036 % \end{macrocode}
6037 % \end{macro}
6038 % \end{macro}
6039 %
6040 % \end{schemadesc}
6041 %
6042 %
6043 % \subsubsection{PDF/* Identification schemata}
6044 % \label{sec:pdfX-id}
6045 %
6046 % \begin{schemadesc}[PDF/A Identification]{\protect\acro{PDF/A} Identification}
6047 % \begin{schemadesc}[PDF/X Identification]{\protect\acro{PDF/X} Identification}
6048 % \begin{schemadesc}[PDF/UA Identification]{\protect\acro{PDF/UA} Identification}
6049 %
6050 % \begin{macro}{\hyxmp@pdfa@id@schema}
6051 % Add properties defined by the \acro{PDF/A} Identification
6052 % schema~\cite{PDFA2008:xmp-props} to the |\hyxmp@xml| macro. These
6053 % properties identify a document as conforming to a particular
6054 % \acro{PDF/A} standard. We default to \PDFstd{A}{1}{b}{} if any
6055 % \acro{PDF/A} compliance is detected but let the author override the
6056 % ``1'' with \optname{pdfapart} and the ``b'' with
6057 % \optname{pdfaconformance}.
6058 % \changes{v2.4}{2013/12/21}{Added this macro}
6059 % \changes{v2.9}{2016/04/12}{Let the author specify the \protect\acro{PDF/A}
6060 % part and conformance IDs, as requested by Leonid Sinev}
6061 % \begin{macrocode}
6062 \newcommand*{\hyxmp@pdfa@id@schema}{%
6063 \ifHy@pdfa
6064 \hyxmp@add@simple{pdfaid:part}{\@pdfapart}%
6065 \hyxmp@add@simple{pdfaid:conformance}{\@pdfaconformance}%
6066 \fi
6067 }
6068 % \end{macrocode}
6069 % \end{macro}
6070 %
6071 % \begin{macro}{\hyxmp@pdfua@id@schema}
6072 % If the document conforms to a \acro{PDF/UA} standard, the author can
6073 % indicate the standard version with \optname{pdfuapart}.
6074 % \changes{v5.0}{2020/02/16}{Added this macro}
6075 % \begin{macrocode}
6076 \newcommand*{\hyxmp@pdfua@id@schema}{%
6077 \hyxmp@add@simple{pdfuaid:part}{\@pdfuapart}%
6078 }
6079 % \end{macrocode}
6080 % \end{macro}
6081 %
6082 % \begin{macro}{\hyxmp@pdfx@id@schema}
6083 % \label{page:pdfx-id-schema}
6084 % If the document conforms to a \acro{PDF/X} standard, the author can
6085 % indicate the standard version with \optname{pdfxstandard}. We
6086 % separately handle \PDFstd{X}{1}{}{}, \PDFstd{X}{2}{}{} and
6087 % \PDFstd{X}{3}{}{}, and \PDFstd{X}{4}{}{} onwards.
6088 % \changes{v5.0}{2020/02/26}{Added this macro}
6089 % \begin{macrocode}
6090 \newcommand*{\hyxmp@pdfx@id@schema}{%
6091 \@hyxmp@count=0\hyxmp@pdfx@major\relax
6092 \ifnum\@hyxmp@count=0
6093 \else
6094 \ifnum\@hyxmp@count=1
6095 \hyxmp@add@simple{pdfx:GTS_PDFXVersion}{PDF/X-1:2001}%
6096 \hyxmp@add@simple{pdfx:GTS_PDFXConformance}{\@pdfxstandard}%
6097 \else
6098 \ifnum\@hyxmp@count<4
6099 \hyxmp@add@simple{pdfx:GTS_PDFXVersion}{\@pdfxstandard}%
6100 \else
6101 \hyxmp@add@simple{pdfxid:GTS_PDFXVersion}{\@pdfxstandard}%
6102 \fi
6103 \fi
6104 \fi
6105 }
6106 % \end{macrocode}
6107 % \end{macro}
6108 %
6109 % \end{schemadesc}
6110 % \end{schemadesc}
6111 % \end{schemadesc}
6112 %
6113 %
6114 % \subsubsection{The IPTC Photo Metadata schema}
6115 % \label{sec:photo-meta}
6116 %
6117 % \begin{schemadesc}[IPTC Photo Metadata]{\protect\acro{IPTC} Photo Metadata}
6118 %
6119 % \begin{macro}{\xmplinesep}
6120 % Lines in multiline fields are separated by |\xmplinesep| in the
6121 % generated \acro{XML}. This defaults to an \acro{LF} (|^^J|) character
6122 % but written as an \acro{XML} character entity for consistency across
6123 % operating systems.
6124 % \changes{v2.2}{2012/12/07}{Added this macro}
6125 % \begin{macrocode}
6126 \begingroup
6127 \catcode`\&=12
6128 \catcode`\#=12
6129 \gdef\xmplinesep{&#xA;}
6130 \endgroup
6131 % \end{macrocode}
6132 % \end{macro}
6133 %
6134 % \begin{macro}{\hyxmp@list@to@lines}
6135 % Given a property~(|#1|) and a macro containing a comma-separated
6136 % list~(|#2|), replace commas with |\xmplinesep|. Do nothing if the
6137 % list is empty.
6138 % \changes{v2.2}{2012/12/07}{Added this macro}
6139 % \begin{macrocode}
6140 \newcommand*{\hyxmp@list@to@lines}[2]{%
6141 \@ifnotmtargexp{#2}{%
6142 \bgroup
6143 \hyxmp@add@to@xml{%
6144 \hyxmp@extra@indent______<#1>%
6145 }%
6146 % \end{macrocode}
6147 % \begin{macro}{\@elt@first}
6148 % The first element of the list is output as is.
6149 % \begin{macrocode}
6150 \def\@elt@first##1{%
6151 \hyxmp@add@to@xml{##1}%
6152 \let\@elt=\@elt@rest
6153 }%
6154 % \end{macrocode}
6155 % \begin{macro}{\@elt@rest}
6156 % The remaining elements of the list are output with a preceding line
6157 % separator (|\xmplinesep|).
6158 % \begin{macrocode}
6159 \def\@elt@rest##1{%
6160 \hyxmp@add@to@xml{\xmplinesep##1}%
6161 }%
6162 % \end{macrocode}
6163 % \begin{macro}{\@elt}
6164 % Re-encode the text from \term{Unicode} if necessary. Then redefine
6165 % |\@elt| to insert a line separator between terms.
6166 % \begin{macrocode}
6167 \let\@elt=\@elt@first
6168 \hyxmp@xmlify{#2}%
6169 \hyxmp@commas@to@list\hyxmp@list{\hyxmp@xmlified}%
6170 \hyxmp@list
6171 \hyxmp@add@to@xml{</#1>^^J}%
6172 \egroup
6173 }%
6174 }
6175 % \end{macrocode}
6176 % \end{macro}
6177 % \end{macro}
6178 % \end{macro}
6179 % \end{macro}
6180 %
6181 % \begin{macro}{\hyxmp@iptc@schema}
6182 % Add properties defined by the \acro{IPTC} Photo Metadata
6183 % schema~\cite{IPTC2010:photo-meta} to the |\hyxmp@xml| macro. We
6184 % currently support only the \xmpprop{Iptc4xmpCore:CreatorContactInfo}
6185 % property, although this is a structure containing multiple fields.
6186 % \changes{v2.2}{2012/12/07}{Added this macro}
6187 % \changes{v2.9}{2016/04/23}{Use \textsf{Iptc4xmpCore} instead of
6188 % \textsf{Iptc4ContInfo} as the contact-information metadata prefix.
6189 % Leonid Sinev reports that Acrobat's \acro{PDF/A} validator seems to
6190 % prefer \textsf{Iptc4xmpCore}}
6191 % \changes{v4.0}{2019/03/09}{Moved the definition of
6192 % \protect\cs{hyxmp@iptc@data} from here into
6193 % \protect\cs{hyxmp@check@iptc@data}}
6194 % \changes{v4.0}{2019/03/16}{Renamed this macro to
6195 % \protect\cs{hyxmp@iptc@schema} from
6196 % \protect\cs{hyxmp@photometa@schema}}
6197 % \changes{v4.0}{2019/03/16}{Rewrote this macro entirely to correct
6198 % the use of fields within a structure}
6199 % \begin{macrocode}
6200 \gdef\hyxmp@iptc@schema{%
6201 % \end{macrocode}
6202 % Because we currently support only
6203 % \xmpprop{Iptc4xmpCore:CreatorContactInfo} it suffices to check if we
6204 % have any relevant data. If so, we instantiate a
6205 % \xmpprop{Iptc4xmpCore:ContactInfo} structure with all available
6206 % fields.
6207 % \begin{macrocode}
6208 \ifx\hyxmp@iptc@data\@empty
6209 \else
6210 \hyxmp@add@to@xml{%
6211 ______<Iptc4xmpCore:CreatorContactInfo rdf:parseType="Resource">^^J%
6212 }%
6213 % \end{macrocode}
6214 % We locally redefine |\hyxmp@extra@indent| to increase the indentation
6215 % of the assignments to \xmpprop{Iptc4xmpCore:CreatorContactInfo}'s
6216 % fields.
6217 % \begin{macrocode}
6218 \bgroup
6219 \edef\hyxmp@extra@indent{\hyxmp@extra@indent\space\space}%
6220 \hyxmp@list@to@lines{Iptc4xmpCore:CiAdrExtadr}{\@pdfcontactaddress}%
6221 \hyxmp@add@simple{Iptc4xmpCore:CiAdrCity}{\@pdfcontactcity}%
6222 \hyxmp@add@simple{Iptc4xmpCore:CiAdrRegion}{\@pdfcontactregion}%
6223 \hyxmp@add@simple{Iptc4xmpCore:CiAdrPcode}{\@pdfcontactpostcode}%
6224 \hyxmp@add@simple{Iptc4xmpCore:CiAdrCtry}{\@pdfcontactcountry}%
6225 % \end{macrocode}
6226 % \begin{macro}{\xmplinesep}
6227 % The \acro{IPTC} standard states that sets of telephone numbers, email
6228 % addresses, and \acro{URL}s for the contact person or institution,
6229 % ``[m]ay have to be separated by a comma in the user
6230 % interface''~\cite{IPTC2010:photo-meta}. This is rather ambiguous:
6231 % Does the comma appear \emph{only} in the user interface or also in the
6232 % generated \acro{XML}\@? Here we assume the latter interpretation and
6233 % temporarily redefine |\xmplinesep| as a comma and use
6234 % |\hyxmp@list@to@lines| to insert the data. Unlike
6235 % |\hyxmp@add@simple|, this approach trims all spaces surrounding commas.
6236 % \begin{macrocode}
6237 \def\xmplinesep{,}%
6238 \hyxmp@list@to@lines{Iptc4xmpCore:CiTelWork}{\@pdfcontactphone}%
6239 \hyxmp@list@to@lines{Iptc4xmpCore:CiEmailWork}{\@pdfcontactemail}%
6240 \hyxmp@list@to@lines{Iptc4xmpCore:CiUrlWork}{\@pdfcontacturl}%
6241 \egroup
6242 \hyxmp@add@to@xml{%
6243 ______</Iptc4xmpCore:CreatorContactInfo>^^J%
6244 }%
6245 \fi
6246 }
6247 % \end{macrocode}
6248 % \end{macro}
6249 % \end{macro}
6250 %
6251 % \end{schemadesc}
6252 %
6253 %
6254 % \subsubsection{The PRISM Basic Metadata schema}
6255 % \label{sec:prism-meta}
6256 %
6257 % \begin{schemadesc}[PRISM Basic Metadata]{\protect\acro{PRISM} Basic Metadata}
6258 %
6259 % \begin{macro}{\hyxmp@prism@schema}
6260 % \changes{v4.0}{2019/03/10}{Added this macro}
6261 % Add properties defined by the \acro{PRISM} Basic Metadata
6262 % schema~\cite{PRISM2012:basic-meta}.
6263 % \begin{macrocode}
6264 \newcommand*{\hyxmp@prism@schema}{%
6265 \ifx\hyxmp@prism@data\@empty
6266 \else
6267 \hyxmp@add@simple{prism:complianceProfile}{three}%
6268 \fi
6269 \hyxmp@add@simple@lang{prism:subtitle}{\@pdfsubtitle}%
6270 \hyxmp@add@simple@lang{prism:publicationName}{\@pdfpublication}%
6271 \hyxmp@add@simple{prism:aggregationType}{\@pdfpubtype}%
6272 \hyxmp@add@simple@lang{prism:bookEdition}{\@pdfbookedition}%
6273 \hyxmp@add@simple{prism:volume}{\@pdfvolumenum}%
6274 \hyxmp@add@simple{prism:number}{\@pdfissuenum}%
6275 \hyxmp@add@simple{prism:pageRange}{\@pdfpagerange}%
6276 \hyxmp@add@simple{prism:isbn}{\@pdfisbn}%
6277 \hyxmp@add@simple{prism:issn}{\@pdfissn}%
6278 \hyxmp@add@simple{prism:eIssn}{\@pdfeissn}%
6279 \hyxmp@add@simple{prism:doi}{\@pdfdoi}%
6280 \hyxmp@add@simple{prism:url}{\@pdfurl}%
6281 \hyxmp@add@simple{prism:byteCount}{\@pdfbytes}%
6282 \hyxmp@add@simple{prism:pageCount}{\@pdfnumpages}%
6283 }
6284 % \end{macrocode}
6285 % \end{macro}
6286 %
6287 % \end{schemadesc}
6288 %
6289 %
6290 % \subsubsection{The Journal Article Versions (JAV) schema}
6291 % \label{sec:niso-jav-meta}
6292 %
6293 % \begin{schemadesc}{Journal Article Versions}
6294 %
6295 % \begin{macro}{\hyxmp@jav@schema}
6296 % \changes{v5.5}{2020/07/23}{Added this macro}
6297 % Add properties defined by the \acro{NISO}/\acro{ALPSP} Journal Article
6298 % Versions schema~\cite{NISO2008:jav}.
6299 % \begin{macrocode}
6300 \newcommand*{\hyxmp@jav@schema}{%
6301 \hyxmp@add@simple{jav:journal_article_version}{\@pdfpubstatus}%
6302 }
6303 % \end{macrocode}
6304 % \end{macro}
6305 %
6306 % \end{schemadesc}
6307 %
6308 % \subsubsection{The XMP Paged-Text schema}
6309 % \label{sec:xmp-paged-text}
6310 %
6311 % \begin{schemadesc}[XMP Paged-Text]{\protect\acro{XMP} Paged-Text}
6312 %
6313 % \begin{macro}{\hyxmp@xmptpg@schema}
6314 % The \acro{XMP} Paged-Text schema~\cite{Adobe2012:XMP} includes
6315 % properties related to the construction of the \acro{PDF} file itself.
6316 % We acquire most of this information through \LuaTeX\ mechanisms and
6317 % therefore include much less information when run from other \TeX\ engines.
6318 % \changes{v5.5}{2020/07/25}{Added this macro}
6319 % \begin{macrocode}
6320 \newcommand*{\hyxmp@xmptpg@schema}{%
6321 \ifLuaTeX
6322 \luadirect{write_xmp_font_list(\the\hyxmp@cct)}%
6323 \fi
6324 \hyxmp@add@simple{xmpTPg:NPages}{\@pdfnumpages}%
6325 }
6326 % \end{macrocode}
6327 % \end{macro}
6328 %
6329 % \begin{macro}{\hyxmp@cct}
6330 % We store the current category-code table to ensure that
6331 % |write_xmp_font_list|'s output uses our redefined category codes.
6332 % \begin{macrocode}
6333 \ifLuaTeX
6334 \newcatcodetable\hyxmp@cct
6335 \savecatcodetable\hyxmp@cct
6336 \fi
6337 % \end{macrocode}
6338 % \end{macro}
6339 %
6340 % \begin{macro}{\hyxmp@prot@us}
6341 % Define an underscore character that's protected from being converted
6342 % into a space when passed to \cs{hyxmp@add@to@xml}. \cs{hyxmp@prot@us}
6343 % is used within |write_xmp_font_list| (below) in particular to typeset
6344 % filenames that may contain underscores.
6345 % \begin{macrocode}
6346 \bgroup
6347 \catcode`\_=11
6348 \gdef\hyxmp@prot@us{_}%
6349 \egroup
6350 % \end{macrocode}
6351 % \end{macro}
6352 %
6353 % Here we define a Lua\index{Lua} function, |write_xmp_font_list|, that
6354 % writes font information to the \acro{XMP} packet.
6355 % \changes{v5.6}{2020/10/02}{Make \protect\texttt{write\_xmp\_font\_list}
6356 % robust to fonts loaded using
6357 % \protect\href{https://en.wikipedia.org/wiki/HarfBuzz}{HarfBuzz}.
6358 % Thanks to John Lienhard for the bug report}
6359 % \changes{v5.6}{2020/10/02}{Don't inadvertently replace underscores in
6360 % filenames when writing font-related metadata}
6361 % \begin{macrocode}
6362 \ifLuaTeX
6363 \begin{luacode*}
6364 function write_xmp_font_list (cct)
6365 local function show_field(name, ...)
6366 for i = 1, select("#", ...) do
6367 local val = select(i, ...)
6368 if val then
6369 local xml = string.gsub(val, "&", "&amp;")
6370 xml = string.gsub(xml, "<", "&lt;")
6371 xml = string.gsub(xml, ">", "&gt;")
6372 xml = string.gsub(xml, "_", "\\hyxmp@prot@us ")
6373 tex.print(cct, "____________<stFnt:" .. name .. ">" ..
6374 xml .. "</stFnt:" .. name .. ">^^J%")
6375 return
6376 end
6377 end
6378 end
6379 tex.print(cct, "\\hyxmp@add@to@xml{%")
6380 tex.print(cct, "______<xmpTPg:Fonts>^^J%")
6381 tex.print(cct, "________<rdf:Bag>^^J%")
6382 for i, f in font.each() do
6383 tex.print(cct, "__________<rdf:li rdf:parseType=\"Resource\">^^J%")
6384 if f.filename then
6385 local fname = string.gsub(f.filename, "^harfloaded:(.*)", "%1")
6386 local info = fontloader.info(fname)
6387 if info then
6388 show_field("fontFace", info.fullname)
6389 show_field("fontFamily", info.familyname)
6390 show_field("fontName", info.fontname)
6391 show_field("versionString", info.version)
6392 end
6393 local baseName = string.gsub(f.filename, ".*[/\\](.*)", "%1")
6394 show_field("fontFileName", baseName)
6395 else
6396 show_field("fontName", f.psname, f.fullname, f.name)
6397 end
6398 if f.format and f.format ~= "unknown" then
6399 show_field("fontType", f.format)
6400 end
6401 tex.print(cct, "__________</rdf:li>^^J%")
6402 end
6403 tex.print(cct, "________</rdf:Bag>^^J%")
6404 tex.print(cct, "______</xmpTPg:Fonts>^^J%")
6405 tex.print(cct, "}")
6406 end
6407 \end{luacode*}
6408 \fi
6409 % \end{macrocode}
6410 %
6411 % \end{schemadesc}
6412 %
6413 %
6414 % \subsubsection{XMP extension schemata}
6415 % \label{sec:extension-schemata}
6416 %
6417 % Not all of the schemata supported by \pkgname{hyperxmp} are predefined
6418 % by \acro{XMP}. \acro{PDF/A} conversion would normally fail for
6419 % documents that employ ``custom'' schemata. However, this problem can
6420 % be circumvented by declaring non-standard schemata in the \acro{XMP}
6421 % packet itself, following a technique described in a PDF Association
6422 % technical note~\cite{PDFA2008:ext-schemas}. In this section, we
6423 % declare only those schemata we actually use.
6424 %
6425 % \begin{macro}{\hyxmp@check@iptc@data}
6426 % \begin{macro}{\hyxmp@iptc@data}
6427 % Define |\hyxmp@iptc@data| as the concatenation of all \acro{IPTC}
6428 % photo metadata supplied by the document.
6429 % \begin{macrocode}
6430 \newcommand*{\hyxmp@check@iptc@data}{%
6431 \edef\hyxmp@iptc@data{%
6432 \@pdfcontactaddress
6433 \@pdfcontactcity
6434 \@pdfcontactregion
6435 \@pdfcontactpostcode
6436 \@pdfcontactcountry
6437 \@pdfcontactphone
6438 \@pdfcontactemail
6439 \@pdfcontacturl
6440 }%
6441 }%
6442 % \end{macrocode}
6443 % \end{macro}
6444 % \end{macro}
6445 %
6446 % \begin{macro}{\hyxmp@check@prism@data}
6447 % \begin{macro}{\hyxmp@prism@data}
6448 % Define |\hyxmp@prism@data| as the concatenation of all \acro{PRISM}
6449 % metadata supplied by the document.
6450 % \begin{macrocode}
6451 \newcommand*{\hyxmp@check@prism@data}{%
6452 \edef\hyxmp@prism@data{%
6453 \@pdfbookedition
6454 \@pdfbytes
6455 \@pdfdoi
6456 \@pdfeissn
6457 \@pdfisbn
6458 \@pdfissn
6459 \@pdfissuenum
6460 \@pdfnumpages
6461 \@pdfpagerange
6462 \@pdfpublication
6463 \@pdfpubtype
6464 \@pdfsubtitle
6465 \@pdfurl
6466 \@pdfvolumenum
6467 }%
6468 }%
6469 % \end{macrocode}
6470 % \end{macro}
6471 % \end{macro}
6472 %
6473 % \begin{macro}{\hyxmp@check@jav@data}
6474 % \begin{macro}{\hyxmp@jav@data}
6475 % Define |\hyxmp@jav@data| as the concatenation of all \acro{JAV}
6476 % metadata supplied by the document.
6477 % \begin{macrocode}
6478 \newcommand*{\hyxmp@check@jav@data}{%
6479 \edef\hyxmp@jav@data{%
6480 \@pdfpubstatus
6481 }%
6482 }
6483 % \end{macrocode}
6484 % \end{macro}
6485 % \end{macro}
6486 %
6487 % \begin{macro}{\hyxmp@begin@extension@decls}
6488 % Begin a block of \acro{XML} tags that indicates we're declaring one or
6489 % more extension schemata.
6490 % \begin{macrocode}
6491 \newcommand*{\hyxmp@begin@extension@decls}{%
6492 \hyxmp@add@to@xml{%
6493 ______<pdfaExtension:schemas>^^J%
6494 ________<rdf:Bag>^^J%
6495 }%
6496 }
6497 % \end{macrocode}
6498 % \end{macro}
6499 %
6500 % \begin{macro}{\hyxmp@end@extension@decls}
6501 % End the block of \acro{XML} tags begun by |\hyxmp@begin@extension@decls|.
6502 % \begin{macrocode}
6503 \newcommand*{\hyxmp@end@extension@decls}{%
6504 \hyxmp@add@to@xml{%
6505 ________</rdf:Bag>^^J%
6506 ______</pdfaExtension:schemas>^^J%
6507 }%
6508 }
6509 % \end{macrocode}
6510 % \end{macro}
6511 %
6512 % \begin{macro}{\hyxmp@begin@ext@decl}
6513 % Begin the declaration of a single extension schema.
6514 % |\hyxmp@begin@ext@decl| accepts the schema's name, prefix, and
6515 % namespace URI.
6516 % \changes{v4.0}{2019/03/15}{Added this macro}
6517 % \begin{macrocode}
6518 \newcommand*{\hyxmp@begin@ext@decl}[3]{%
6519 \hyxmp@add@to@xml{%
6520 __________<rdf:li rdf:parseType="Resource">^^J%
6521 ____________<pdfaSchema:schema>#1</pdfaSchema:schema>^^J%
6522 ____________<pdfaSchema:prefix>#2</pdfaSchema:prefix>^^J%
6523 ____________<pdfaSchema:namespaceURI>#3</pdfaSchema:namespaceURI>^^J%
6524 ____________<pdfaSchema:property>^^J%
6525 ______________<rdf:Seq>^^J%
6526 }%
6527 }%
6528 % \end{macrocode}
6529 % \end{macro}
6530 %
6531 % \begin{macro}{\hyxmp@end@ext@decl}
6532 % End the declaration of a single extension schema.
6533 % \changes{v4.0}{2019/03/15}{Added this macro}
6534 % \begin{macrocode}
6535 \newcommand*{\hyxmp@end@ext@decl}{%
6536 \hyxmp@add@to@xml{%
6537 ______________</rdf:Seq>^^J%
6538 ____________</pdfaSchema:property>^^J%
6539 __________</rdf:li>^^J%
6540 }%
6541 }%
6542 % \end{macrocode}
6543 % \end{macro}
6544 %
6545 % \begin{macro}{\hyxmp@declare@property}
6546 % Declare a single extension-schema property. |\hyxmp@declare@property|
6547 % takes as input an optional type (defaults to \xmpterm{Text}) and a
6548 % mandatory name, category, and description.
6549 % \changes{v4.0}{2019/03/15}{Added this macro}
6550 % \changes{v5.0}{2020/02/21}{Insert the property name (\string\texttt{\#2})
6551 % verbatim}
6552 % \begin{macrocode}
6553 \newcommand{\hyxmp@declare@property}[4][Text]{%
6554 \hyxmp@add@to@xml{%
6555 ________________<rdf:li rdf:parseType="Resource">^^J%
6556 __________________<pdfaProperty:name>}%
6557 \xdef\hyxmp@xml{\hyxmp@xml#2}%
6558 \hyxmp@add@to@xml{</pdfaProperty:name>^^J%
6559 __________________<pdfaProperty:valueType>#1</pdfaProperty:valueType>^^J%
6560 __________________<pdfaProperty:category>#3</pdfaProperty:category>^^J%
6561 __________________<pdfaProperty:description>#4</pdfaProperty:description>^^J%
6562 ________________</rdf:li>^^J%
6563 }%
6564 }%
6565 % \end{macrocode}
6566 % \end{macro}
6567 %
6568 % \begin{macro}{\hyxmp@declare@field}
6569 % Declare a single field in a custom datatype required by an extension
6570 % schema. |\hyxmp@declare@field| takes as input an optional type
6571 % (defaults to \xmpterm{Text}) and a mandatory name and description.
6572 % \changes{v4.0}{2019/03/16}{Replaced
6573 % \protect\cs{hyxmp@declare@resource} with this macro}
6574 % \begin{macrocode}
6575 \newcommand{\hyxmp@declare@field}[3][Text]{%
6576 \hyxmp@add@to@xml{%
6577 ______________________<rdf:li rdf:parseType="Resource">^^J%
6578 ________________________<pdfaField:name>#2</pdfaField:name>^^J%
6579 ________________________<pdfaField:valueType>#1</pdfaField:valueType>^^J%
6580 ________________________<pdfaField:description>#3</pdfaField:description>^^J%
6581 ______________________</rdf:li>^^J%
6582 }%
6583 }
6584 % \end{macrocode}
6585 % \end{macro}
6586 %
6587 % \begin{macro}{\hyxmp@pdf@extensions}
6588 % Declare the Adobe \acro{PDF} schema.
6589 % \changes{v5.0}{2020/02/27}{Added this macro}
6590 % \begin{macrocode}
6591 \newcommand*{\hyxmp@pdf@extensions}{%
6592 \hyxmp@begin@ext@decl
6593 {Adobe PDF Schema}%
6594 {pdf}%
6595 {http://ns.adobe.com/pdf/1.3/}%
6596 \hyxmp@declare@property
6597 {Trapped}%
6598 {internal}%
6599 {Indication if the document has been modified to include trapping information}%
6600 \hyxmp@end@ext@decl
6601 }%
6602 % \end{macrocode}
6603 % \end{macro}
6604 %
6605 % \begin{macro}{\hyxmp@mm@extensions}
6606 % Declare the \acro{XMP} Media Management schema.
6607 % \changes{v4.0}{2019/03/15}{Added this macro}
6608 % \changes{v5.5}{2020/07/23}{Corrected the type of
6609 % \protect\xmpprop{xmpMM:RenditionClass}. Thanks to Thorsten Wi{\ss}mann
6610 % for the bug report and patch}
6611 % \begin{macrocode}
6612 \newcommand*{\hyxmp@mm@extensions}{%
6613 \hyxmp@begin@ext@decl
6614 {XMP Media Management Schema}%
6615 {xmpMM}%
6616 {http://ns.adobe.com/xap/1.0/mm/}%
6617 \hyxmp@declare@property
6618 [URI]
6619 {DocumentID}%
6620 {internal}%
6621 {UUID based identifier for all versions and renditions of a document}%
6622 \hyxmp@declare@property
6623 [URI]
6624 {InstanceID}%
6625 {internal}%
6626 {UUID based identifier for specific incarnation of a document}%
6627 \hyxmp@declare@property
6628 {VersionID}%
6629 {internal}%
6630 {Document version identifier}%
6631 \hyxmp@declare@property
6632 [RenditionClass]%
6633 {RenditionClass}%
6634 {internal}%
6635 {The manner in which a document is rendered}%
6636 \hyxmp@end@ext@decl
6637 }%
6638 % \end{macrocode}
6639 % \end{macro}
6640 %
6641 % \begin{macro}{\hyxmp@pdfa@id@extensions}
6642 % Declare the \acro{PDF/A} Identification schema~\cite{PDFA2008:xmp-props}.
6643 % \changes{v4.0}{2019/03/15}{Added this macro}
6644 % \begin{macrocode}
6645 \newcommand*{\hyxmp@pdfa@id@extensions}{%
6646 \hyxmp@begin@ext@decl
6647 {PDF/A Identification Schema}%
6648 {pdfaid}%
6649 {http://www.aiim.org/pdfa/ns/id/}%
6650 \hyxmp@declare@property
6651 [Integer]%
6652 {part}%
6653 {internal}%
6654 {Part of PDF/A standard}%
6655 \hyxmp@declare@property
6656 {conformance}%
6657 {internal}%
6658 {Conformance level of PDF/A standard}%
6659 \hyxmp@end@ext@decl
6660 }%
6661 % \end{macrocode}
6662 % \end{macro}
6663 %
6664 % \begin{macro}{\hyxmp@pdfua@id@extensions}
6665 % Declare the \acro{PDF/UA} Universal Accessibility schema.
6666 % \changes{v5.0}{2020/02/15}{Added this macro}
6667 % \begin{macrocode}
6668 \newcommand*{\hyxmp@pdfua@id@extensions}{%
6669 \hyxmp@begin@ext@decl
6670 {PDF/UA Universal Accessibility Schema}%
6671 {pdfuaid}%
6672 {http://www.aiim.org/pdfua/ns/id/}%
6673 \hyxmp@declare@property
6674 [Integer]%
6675 {part}%
6676 {internal}%
6677 {Part of ISO 14289 standard}%
6678 \hyxmp@end@ext@decl
6679 }%
6680 % \end{macrocode}
6681 % \end{macro}
6682 %
6683 % \begin{macro}{\hyxmp@pdfx@id@extensions}
6684 % Declare the schema used pre-\PDFstd{X}{4}{}{}. Because Adobe Acrobat DC
6685 % (at least) defines this even for \PDFstd{X}{4}{}{} and later, we follow
6686 % suit.
6687 % \changes{v5.0}{2020/02/26}{Added this macro}
6688 % \begin{macrocode}
6689 \newcommand*{\hyxmp@pdfx@id@extensions}{%
6690 \ifx\hyxmp@pdfx@major\empty
6691 \else
6692 \hyxmp@begin@ext@decl
6693 {Adobe Document Info PDF/X eXtension Schema}%
6694 {pdfx}%
6695 {http://ns.adobe.com/pdfx/1.3/}%
6696 \hyxmp@declare@property
6697 {GTS_PDFXVersion}%
6698 {internal}%
6699 {ID of PDF/X standard}%
6700 \hyxmp@declare@property
6701 {GTS_PDFXConformance}%
6702 {internal}%
6703 {Conformance level of PDF/X standard}%
6704 \hyxmp@end@ext@decl
6705 \fi
6706 % \end{macrocode}
6707 % Declare the schema used in \PDFstd{X}{4}{}{} and later versions.
6708 % \begin{macrocode}
6709 \@hyxmp@count=0\hyxmp@pdfx@major\relax
6710 \ifnum\@hyxmp@count>3
6711 \hyxmp@begin@ext@decl
6712 {PDF/X ID Schema}%
6713 {pdfxid}%
6714 {http://www.npes.org/pdfx/ns/id/}%
6715 \hyxmp@declare@property
6716 {GTS_PDFXVersion}%
6717 {internal}%
6718 {ID of PDF/X standard}%
6719 \hyxmp@end@ext@decl
6720 \fi
6721 }%
6722 % \end{macrocode}
6723 % \end{macro}
6724 %
6725 % \begin{macro}{\hyxmp@iptc@extensions}
6726 % \changes{v2.2}{2012/12/13}{Added this macro to support
6727 % \acro{PDF/A} generation}
6728 % \changes{v2.3}{2013/01/08}{Gave the
6729 % \xmpprop{Iptc4xmpCore:CreatorContactInfo} fields a unique
6730 % \xmpprop{pdfaType:prefix} to better support conversion of the
6731 % document to \acro{PDF/A}}
6732 % \changes{v4.0}{2019/03/09}{Moved the header code from here into
6733 % \protect\cs{hyxmp@begin@extension@decls} and the trailer code
6734 % from here into \protect\cs{hyxmp@end@extension@decls}}
6735 % \changes{v4.0}{2019/03/16}{Rewrote to more closely honor the
6736 % \acro{XMP} specification}
6737 % Because \acro{IPTC} metadata are not recognized by the \acro{PDF/A}
6738 % standard, \acro{PDF/A} conversion would normally fail for documents
6739 % that utilize \acro{IPTC} metadata. Declaring the \acro{IPTC}
6740 % metadata we support enables the document to be converted to
6741 % \acro{PDF/A} format.
6742 % \begin{macrocode}
6743 \newcommand*{\hyxmp@iptc@extensions}{%
6744 \hyxmp@begin@ext@decl
6745 {IPTC Core Schema}%
6746 {Iptc4xmpCore}%
6747 {http://iptc.org/std/Iptc4xmpCore/1.0/xmlns/}%
6748 \hyxmp@declare@property
6749 [ContactInfo]
6750 {CreatorContactInfo}
6751 {external}
6752 {Document creator's contact information}
6753 % \end{macrocode}
6754 % We can't call |\hyxmp@end@ext@decl| because we need first need to
6755 % define the \xmpprop{Iptc4xmpCore:ContactInfo} structure.
6756 % \begin{macrocode}
6757 \hyxmp@add@to@xml{%
6758 ______________</rdf:Seq>^^J%
6759 ____________</pdfaSchema:property>^^J%
6760 ____________<pdfaSchema:valueType>^^J%
6761 ______________<rdf:Seq>^^J%
6762 ________________<rdf:li rdf:parseType="Resource">^^J%
6763 __________________<pdfaType:type>ContactInfo</pdfaType:type>^^J%
6764 __________________<pdfaType:namespaceURI>http://iptc.org/std/Iptc4xmpCore/1.0/xmlns/</pdfaType:namespaceURI>^^J%
6765 __________________<pdfaType:prefix>Iptc4xmpCore</pdfaType:prefix>^^J%
6766 __________________<pdfaType:description>%
6767 Basic set of information to get in contact with a person%
6768 </pdfaType:description>^^J%
6769 __________________<pdfaType:field>^^J%
6770 ____________________<rdf:Seq>^^J%
6771 }%
6772 \hyxmp@declare@field
6773 {CiAdrCity}%
6774 {Contact information city}%
6775 \hyxmp@declare@field
6776 {CiAdrCtry}%
6777 {Contact information country}%
6778 \hyxmp@declare@field
6779 {CiAdrExtadr}%
6780 {Contact information address}%
6781 \hyxmp@declare@field
6782 {CiAdrPcode}%
6783 {Contact information local postal code}%
6784 \hyxmp@declare@field
6785 {CiAdrRegion}%
6786 {Contact information regional information such as state or province}%
6787 \hyxmp@declare@field
6788 {CiEmailWork}%
6789 {Contact information email address(es)}%
6790 \hyxmp@declare@field
6791 {CiTelWork}%
6792 {Contact information telephone number(s)}%
6793 \hyxmp@declare@field
6794 {CiUrlWork}%
6795 {Contact information Web URL(s)}%
6796 \hyxmp@add@to@xml{%
6797 ____________________</rdf:Seq>^^J%
6798 __________________</pdfaType:field>^^J%
6799 ________________</rdf:li>^^J%
6800 ______________</rdf:Seq>^^J%
6801 ____________</pdfaSchema:valueType>^^J%
6802 __________</rdf:li>^^J%
6803 }%
6804 }
6805 % \end{macrocode}
6806 % \end{macro}
6807 %
6808 % \begin{macro}{\hyxmp@prism@extensions}
6809 % \changes{v4.0}{2019/03/15}{Added this macro}
6810 % Because \acro{PRISM} metadata are not recognized by the \acro{PDF/A}
6811 % standard, \acro{PDF/A} conversion would normally fail for documents
6812 % that utilize \acro{PRISM} metadata. Declaring the \acro{PRISM}
6813 % metadata we support enables the document to be converted to
6814 % \acro{PDF/A} format.
6815 % \begin{macrocode}
6816 \newcommand*{\hyxmp@prism@extensions}{%
6817 \hyxmp@begin@ext@decl
6818 {PRISM Basic Metadata}%
6819 {prism}%
6820 {http://prismstandard.org/namespaces/basic/3.0/}%
6821 \hyxmp@declare@property
6822 {complianceProfile}%
6823 {internal}%
6824 {PRISM specification compliance profile to which this document adheres}%
6825 \hyxmp@declare@property
6826 {publicationName}%
6827 {external}%
6828 {Publication name}%
6829 \hyxmp@declare@property
6830 {aggregationType}%
6831 {external}%
6832 {Publication type}%
6833 \hyxmp@declare@property
6834 {bookEdition}%
6835 {external}%
6836 {Edition of the book in which the document was published}%
6837 \hyxmp@declare@property
6838 {volume}%
6839 {external}%
6840 {Publication volume number}%
6841 \hyxmp@declare@property
6842 {number}%
6843 {external}%
6844 {Publication issue number within a volume}%
6845 \hyxmp@declare@property
6846 {pageRange}%
6847 {external}%
6848 {Page range for the document within the print version of its publication}%
6849 \hyxmp@declare@property
6850 {issn}%
6851 {external}%
6852 {ISSN for the printed publication in which the document was published}%
6853 \hyxmp@declare@property
6854 {eIssn}%
6855 {external}%
6856 {ISSN for the electronic publication in which the document was published}%
6857 \hyxmp@declare@property
6858 {isbn}%
6859 {external}%
6860 {ISBN for the publication in which the document was published}%
6861 \hyxmp@declare@property
6862 {doi}%
6863 {external}%
6864 {Digital Object Identifier for the document}%
6865 \hyxmp@declare@property
6866 [URL]
6867 {url}%
6868 {external}%
6869 {URL at which the document can be found}%
6870 \hyxmp@declare@property
6871 [Integer]
6872 {byteCount}%
6873 {internal}%
6874 {Approximate file size in octets}%
6875 \hyxmp@declare@property
6876 [Integer]
6877 {pageCount}%
6878 {internal}%
6879 {Number of pages in the print version of the document}%
6880 \hyxmp@declare@property
6881 {subtitle}%
6882 {external}%
6883 {Document's subtitle}%
6884 \hyxmp@end@ext@decl
6885 }%
6886 % \end{macrocode}
6887 % \end{macro}
6888 %
6889 % \begin{macro}{\hyxmp@jav@extensions}
6890 % \changes{v5.5}{2020/07/23}{Added this macro}
6891 % Because \acro{JAV} metadata are not recognized by the \acro{PDF/A}
6892 % standard, \acro{PDF/A} conversion would normally fail for documents
6893 % that utilize \acro{JAV} metadata. Declaring the \acro{JAV} metadata
6894 % we support enables the document to be converted to \acro{PDF/A}
6895 % format.
6896 % \begin{macrocode}
6897 \newcommand*{\hyxmp@jav@extensions}{%
6898 \hyxmp@begin@ext@decl
6899 {NISO/ALPSP Journal Article Versions}%
6900 {jav}%
6901 {http://www.niso.org/schemas/jav/1.0/}%
6902 \hyxmp@declare@property
6903 [Closed Choice of Text]%
6904 {journal_article_version}%
6905 {external}%
6906 {Article status, one of "AO", "SMUR", "AM", "P", "VoR", "CVoR", or "EVoR"}%
6907 \hyxmp@end@ext@decl
6908 }%
6909 % \end{macrocode}
6910 % \end{macro}
6911 %
6912 % \begin{macro}{\hyxmp@declare@extensions}
6913 % Declare all \acro{XMP} extension schemata. We'll always have at least
6914 % one, the \acro{XMP} Media Management extensions, because we
6915 % automatically generate \xmpprop{xmpMM:DocumentID} and
6916 % \xmpprop{xmpMM:InstanceID}
6917 % values.
6918 % \begin{macrocode}
6919 \newcommand*{\hyxmp@declare@extensions}{%
6920 \hyxmp@begin@extension@decls
6921 % \end{macrocode}
6922 % Declare the Adobe \acro{PDF} schema (always present).
6923 % \begin{macrocode}
6924 \hyxmp@pdf@extensions
6925 % \end{macrocode}
6926 % Declare the \acro{XMP} Media Management extensions (always present).
6927 % \begin{macrocode}
6928 \hyxmp@mm@extensions
6929 % \end{macrocode}
6930 % Declare the \acro{PDF/A} Identification extensions, but only when
6931 % generating a \acro{PDF/A} document.
6932 % \begin{macrocode}
6933 \ifHy@pdfa
6934 \hyxmp@pdfa@id@extensions
6935 \fi
6936 % \end{macrocode}
6937 % Conditionally declare the \acro{PDF/UA} Universal Accessibility
6938 % extensions.
6939 % \begin{macrocode}
6940 \ifx\@pdfuapart\@empty
6941 \else
6942 \hyxmp@pdfua@id@extensions
6943 \fi
6944 % \end{macrocode}
6945 % \begin{macro}{\next}
6946 % Conditionally declare the \acro{PDF/X} extensions.
6947 % \begin{macrocode}
6948 \ifx\@pdfxversion\@empty
6949 \else
6950 \hyxmp@pdfx@id@extensions
6951 \fi
6952 % \end{macrocode}
6953 % \end{macro}
6954 % Conditionally declare \acro{IPTC} photo metadata extensions.
6955 % \begin{macrocode}
6956 \ifx\hyxmp@iptc@data\@empty
6957 \else
6958 \hyxmp@iptc@extensions
6959 \fi
6960 % \end{macrocode}
6961 % Conditionally declare \acro{PRISM} basic metadata extensions.
6962 % \begin{macrocode}
6963 \ifx\hyxmp@prism@data\@empty
6964 \else
6965 \hyxmp@prism@extensions
6966 \fi
6967 % \end{macrocode}
6968 % Conditionally declare \acro{JAV} metadata.
6969 % \begin{macrocode}
6970 \ifx\hyxmp@jav@data\@empty
6971 \else
6972 \hyxmp@jav@extensions
6973 \fi
6974 % \end{macrocode}
6975 % \begin{macrocode}
6976 \hyxmp@end@extension@decls
6977 }
6978 % \end{macrocode}
6979 % \end{macro}
6980 %
6981 % \subsubsection{Combining schemata into an XMP packet}
6982 % \label{sec:combining-schemata}
6983 %
6984 % \begin{macro}{\hyxmp@bom}
6985 % Define a macro for the \term{Unicode} byte-order marker (\acro{BOM}).
6986 % \changes{v2.0}{2012/08/02}{Added by Heiko Oberdiek}
6987 % \begin{macrocode}
6988 \begingroup
6989 \ifhyxmp@unicodetex
6990 \lccode`\!="FEFF %
6991 \lowercase{%
6992 \gdef\hyxmp@bom{!}
6993 }%
6994 \else
6995 \catcode`\^^ef=12
6996 \catcode`\^^bb=12
6997 \catcode`\^^bf=12
6998 \gdef\hyxmp@bom{^^ef^^bb^^bf}%
6999 \fi
7000 \endgroup
7001 % \end{macrocode}
7002 % \end{macro}
7003 %
7004 % \begin{macro}{\hyxmp@construct@packet}
7005 % \changes{v1.1}{2006/05/21}{Explicitly set the category codes of
7006 % characters \string\meta{EF}, \string\meta{BB}, and
7007 % \string\meta{BF} to ``letter''. Thanks to Daniel Sch\"omer for
7008 % the bug report}
7009 % \changes{v2.0}{2012/08/02}{Modified by Heiko Oberdiek to use an
7010 % appropriate \acro{BOM} representation via
7011 % \texttt{\string\string\string\hyxmp@bom}}
7012 % \begin{macro}{\hyxmp@xml}
7013 % Successively add \acro{XML} data to |\hyxmp@xml| until we have
7014 % something we can insert into the document's \acro{PDF} catalog.
7015 % \begin{macrocode}
7016 \def\hyxmp@construct@packet{%
7017 \gdef\hyxmp@xml{}%
7018 \hyxmp@add@to@xml{<?xpacket begin="\hyxmp@bom" %
7019 id="W5M0MpCehiHzreSzNTczkc9d"?>^^J%
7020 <x:xmpmeta xmlns:x="adobe:ns:meta/">^^J%
7021 __<rdf:RDF %
7022 xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns\hyxmp@hash">^^J%
7023 ____<rdf:Description rdf:about=""^^J%
7024 % \end{macrocode}
7025 % Specify every namespace we can potentially use, even the ones we end
7026 % up not actually using.
7027 % \begin{macrocode}
7028 _____________________xmlns:pdf="http://ns.adobe.com/pdf/1.3/"^^J%
7029 _____________________xmlns:xmpRights="http://ns.adobe.com/xap/1.0/rights/"^^J%
7030 _____________________xmlns:dc="http://purl.org/dc/elements/1.1/"^^J%
7031 _____________________xmlns:photoshop="http://ns.adobe.com/photoshop/1.0/"^^J%
7032 _____________________xmlns:xmp="http://ns.adobe.com/xap/1.0/"^^J%
7033 _____________________xmlns:xmpMM="http://ns.adobe.com/xap/1.0/mm/"^^J%
7034 _____________________xmlns:stEvt="http://ns.adobe.com/xap/1.0/sType/ResourceEvent\hyxmp@hash"^^J%
7035 _____________________xmlns:pdfaid="http://www.aiim.org/pdfa/ns/id/"^^J%
7036 _____________________xmlns:pdfuaid="http://www.aiim.org/pdfua/ns/id/"^^J%
7037 _____________________xmlns:pdfx="http://ns.adobe.com/pdfx/1.3/"^^J%
7038 }%
7039 % \end{macrocode}
7040 % We make one exception to the rule of including every namespace we can
7041 % potentially use: We don't define the \texttt{pdfxid} namespace unless
7042 % the \acro{PDF/X} version (specified by the \optname{pdfxstandard})
7043 % option is~4 or greater. Otherwise, Adobe Acrobat---at least Adobe
7044 % Acrobat DC 2020---alters the way it displays color. (I believe it
7045 % renders color in a printer gamut instead of a screen gamut.)
7046 % \changes{v5.10}{2020/01/05}{Include the \string\texttt{pdfxid} namespace
7047 % only if the \protect\acro{PDF/X} version is~4 or greater. Thanks to
7048 % John Lienhard for the bug report}
7049 % \begin{macrocode}
7050 \ifnum0\hyxmp@pdfx@major>3
7051 \hyxmp@add@to@xml{%
7052 _____________________xmlns:pdfxid="http://www.npes.org/pdfx/ns/id/"^^J%
7053 }%
7054 \fi
7055 % \end{macrocode}
7056 % Revert to ``include every namespace'' mode.
7057 % \begin{macrocode}
7058 \hyxmp@add@to@xml{%
7059 _____________________xmlns:prism="http://prismstandard.org/namespaces/basic/3.0/"^^J%
7060 _____________________xmlns:jav="http://www.niso.org/schemas/jav/1.0/"^^J%
7061 _____________________xmlns:xmpTPg="http://ns.adobe.com/xap/1.0/t/pg/"^^J%
7062 _____________________xmlns:stFnt="http://ns.adobe.com/xap/1.0/sType/Font\hyxmp@hash"^^J%
7063 _____________________xmlns:Iptc4xmpCore="http://iptc.org/std/Iptc4xmpCore/1.0/xmlns/"^^J%
7064 _____________________xmlns:pdfaExtension="http://www.aiim.org/pdfa/ns/extension/"^^J%
7065 _____________________xmlns:pdfaSchema="http://www.aiim.org/pdfa/ns/schema\hyxmp@hash"^^J%
7066 _____________________xmlns:pdfaProperty="http://www.aiim.org/pdfa/ns/property\hyxmp@hash"^^J%
7067 _____________________xmlns:pdfaType="http://www.aiim.org/pdfa/ns/type\hyxmp@hash"^^J%
7068 _____________________xmlns:pdfaField="http://www.aiim.org/pdfa/ns/field\hyxmp@hash">^^J%
7069 }%
7070 % \end{macrocode}
7071 % Declare non-standard schemata.
7072 % \begin{macrocode}
7073 \hyxmp@check@iptc@data
7074 \hyxmp@check@prism@data
7075 \hyxmp@check@jav@data
7076 \hyxmp@declare@extensions
7077 % \end{macrocode}
7078 % Insert all the metadata we know how to insert.
7079 % \begin{macrocode}
7080 \hyxmp@pdf@schema
7081 \hyxmp@xmpRights@schema
7082 \hyxmp@dc@schema
7083 \hyxmp@photoshop@schema
7084 \hyxmp@xmp@basic@schema
7085 \hyxmp@pdfa@id@schema
7086 \hyxmp@pdfua@id@schema
7087 \hyxmp@pdfx@id@schema
7088 \hyxmp@mm@schema
7089 \hyxmp@iptc@schema
7090 \hyxmp@prism@schema
7091 \hyxmp@jav@schema
7092 \hyxmp@xmptpg@schema
7093 \hyxmp@add@to@xml{%
7094 ____</rdf:Description>^^J%
7095 __</rdf:RDF>^^J%
7096 </x:xmpmeta>^^J%
7097 \hyxmp@padding
7098 <?xpacket end="w"?>^^J%
7099 }%
7100 }
7101 % \end{macrocode}
7102 % \end{macro}
7103 % \end{macro}
7104 %
7105 %
7106 % \subsection{Embedding the XMP packet}
7107 %
7108 % The \acro{PDF} specification says that ``a metadata stream may be
7109 % attached to a document through the \pdfterm{Metadata} entry in the
7110 % document catalogue''~\cite{Adobe2008:PDF} so that's what we do here.
7111 %
7112 % \begin{macro}{\hyxmp@embed@packet}
7113 % \begin{macro}{\hyxmp@driver}
7114 % Determine which \pkgname{hyperref} driver is in use and invoke the
7115 % appropriate embedding function.
7116 % \begin{macrocode}
7117 \newcommand*{\hyxmp@embed@packet}{%
7118 \hyxmp@construct@packet
7119 \def\hyxmp@driver{hpdftex}%
7120 \ifx\hyxmp@driver\Hy@driver
7121 \hyxmp@embed@packet@pdftex
7122 \else
7123 \def\hyxmp@driver{hluatex}%
7124 \ifx\hyxmp@driver\Hy@driver
7125 \hyxmp@embed@packet@luatex
7126 \else
7127 \def\hyxmp@driver{hdvipdfm}%
7128 \ifx\hyxmp@driver\Hy@driver
7129 \hyxmp@embed@packet@dvipdfm
7130 \else
7131 \def\hyxmp@driver{hxetex}%
7132 \ifx\hyxmp@driver\Hy@driver
7133 \hyxmp@embed@packet@xetex
7134 \else
7135 \@ifundefined{pdfmark}{%
7136 \PackageWarningNoLine{hyperxmp}{%
7137 Unrecognized hyperref driver `\Hy@driver'.\MessageBreak
7138 \[email protected]'s XMP metadata will *not* be\MessageBreak
7139 embedded in the resulting file}%
7140 }{%
7141 \hyxmp@embed@packet@pdfmark
7142 }%
7143 \fi
7144 \fi
7145 \fi
7146 \fi
7147 }
7148 % \end{macrocode}
7149 % \end{macro}
7150 % \end{macro}
7151 %
7152 % \subsubsection{Embedding using \pdfTeX}
7153 %
7154 % Up to version~0.85, \LuaTeX\ supported the \pdfTeX\ primitives, and
7155 % \pkgname{hyperref} didn't distinguish the two backends. However, from
7156 % \pkgname{hyperxmp}'s perspective there is one key difference: the
7157 % effect of |\pdfcompresslevel| is local to a group in \pdfTeX\ but is
7158 % global in \LuaTeX\@.
7159 %
7160 % The \acro{PDF} object representing the \acro{XMP} packet is supposed
7161 % to include an uncompressed stream so it can be read by
7162 % non-\acro{PDF}-aware tools. However, we don't want to unnecessarily
7163 % uncompress \emph{every} \acro{PDF} stream. The solution, provided by
7164 % Hans Hagen on the |luatex| mailing list (thread:
7165 % \href{http://tug.org/pipermail/luatex/2016-July/006077.html}{``Leaving
7166 % a single \acro{PDF} object uncompressed''}, 6\,\textsc{Jul}\,2016),
7167 % is to provide the |uncompressed| flag to |\pdfobj|. Our definition of
7168 % |\hyxmp@embed@packet@pdftex| uses the \pkgname{ifluatex} package to
7169 % distinguish the \pdfTeX\ case from the pre-0.85 \LuaTeX\ case.
7170 %
7171 % \begin{macrocode}
7172 \RequirePackage{ifluatex}
7173 % \end{macrocode}
7174 %
7175 % \begin{macro}{\hyxmp@embed@packet@pdftex}
7176 % Embed the \acro{XMP} packet using \pdfTeX\ primitives, which are
7177 % supported by both \pdfTeX\ and pre-0.85 \LuaTeX. The only difference
7178 % is that in the former case we locally specify |\pdfcompresslevel=0| to
7179 % leave the \acro{PDF} object uncompressed while in the latter case we
7180 % pass the |uncompressed| flag to |\pdfobj| to achieve the same effect.
7181 % \changes{v3.1}{2016/07/06}{Leave the \protect\acro{XMP} packet---and
7182 % only the \protect\acro{XMP} packet---uncompressed in both
7183 % \protect\pdfTeX\ and pre-0.85 \protect\LuaTeX}
7184 % \begin{macrocode}
7185 \newcommand*{\hyxmp@embed@packet@pdftex}{%
7186 \bgroup
7187 \ifluatex
7188 \else
7189 \pdfcompresslevel=0
7190 \fi
7191 \immediate\pdfobj \ifluatex uncompressed\fi stream attr {%
7192 /Type /Metadata
7193 /Subtype /XML
7194 }{\hyxmp@xml}%
7195 \pdfcatalog {/Metadata \the\pdflastobj\space 0 R}%
7196 \egroup
7197 }
7198 % \end{macrocode}
7199 % \end{macro}
7200 %
7201 % \subsubsection{Embedding using \LuaTeX~0.85+}
7202 %
7203 % \begin{macro}{\hyxmp@embed@packet@luatex}
7204 % Embed the \acro{XMP} packet using \LuaTeX~0.85+ primitives.
7205 % \changes{v3.0}{2016/07/02}{Added this macro}
7206 % \changes{v3.1}{2016/07/06}{Updated to use
7207 % \protect\texttt{\string\string\string\pdfextension\ obj uncompressed}
7208 % as suggested by Hans Hagen}
7209 % \begin{macrocode}
7210 \newcommand*{\hyxmp@embed@packet@luatex}{%
7211 \immediate\pdfextension obj uncompressed stream attr {%
7212 /Type /Metadata
7213 /Subtype /XML
7214 }{\hyxmp@xml}%
7215 \pdfextension catalog {/Metadata \the\numexpr\pdffeedback lastobj\relax\space 0 R}%
7216 }
7217 % \end{macrocode}
7218 % \end{macro}
7219 %
7220 % \subsubsection{Embedding using any \texttt{pdfmark}-based backend}
7221 %
7222 % \begin{macro}{\hyxmp@embed@packet@pdfmark}
7223 % Embed the \acro{XMP} packet using \pkgname{hyperref}'s |\pdfmark|
7224 % command. I believe |\pdfmark| is used by the \optname{dvipdf},
7225 % \optname{dvipsone}, \optname{dvips}, \optname{dviwindo},
7226 % \optname{nativepdf}, \optname{pdfmark}, \optname{ps2pdf},
7227 % \optname{textures}, and \optname{vtexpdfmark} options to
7228 % \pkgname{hyperref}, but I've tested only a few of those.
7229 % \begin{macrocode}
7230 \newcommand*{\hyxmp@embed@packet@pdfmark}{%
7231 \pdfmark{%
7232 pdfmark=/NamespacePush
7233 }%
7234 \pdfmark{%
7235 pdfmark=/OBJ,
7236 Raw={/_objdef \string{hyxmp@Metadata\string} /type /stream}%
7237 }%
7238 \pdfmark{%
7239 pdfmark=/PUT,
7240 Raw={\string{hyxmp@Metadata\string}
7241 2 dict begin
7242 /Type /Metadata def
7243 /Subtype /XML def
7244 currentdict
7245 end
7246 }%
7247 }%
7248 \pdfmark{%
7249 pdfmark=/PUT,
7250 Raw={\string{hyxmp@Metadata\string} (\hyxmp@xml)}%
7251 }%
7252 \pdfmark{%
7253 pdfmark=/Metadata,
7254 Raw={\string{Catalog\string} \string{hyxmp@Metadata\string}}%
7255 }%
7256 \pdfmark{%
7257 pdfmark=/NamespacePop
7258 }%
7259 }
7260 % \end{macrocode}
7261 % \end{macro}
7262 %
7263 % \subsubsection{Embedding using \texttt{dvipdfm}}
7264 %
7265 % \begin{macro}{\hyxmp@embed@packet@dvipdfm}
7266 % Embed the \acro{XMP} packet using \cmdname{dvipdfm}-specific
7267 % |\special| commands. Note that \cmdname{dvipdfm} rather irritatingly
7268 % requires us to count the number of characters in the |\hyxmp@xml|
7269 % stream ourselves.
7270 % \begin{macrocode}
7271 \newcommand*{\hyxmp@embed@packet@dvipdfm}{%
7272 \hyxmp@string@len{\hyxmp@xml}%
7273 \special{pdf: object @hyxmp@Metadata
7274 <<
7275 /Type /Metadata
7276 /Subtype /XML
7277 /Length \the\@hyxmp@count
7278 >>
7279 stream^^J\hyxmp@xml endstream%
7280 }%
7281 \special{pdf: docview
7282 <<
7283 /Metadata @hyxmp@Metadata
7284 >>
7285 }%
7286 }
7287 % \end{macrocode}
7288 % \end{macro}
7289 %
7290 % \begin{macro}{\hyxmp@string@len}
7291 % Set |\@hyxmp@count| to the number of characters in a given string~(|#1|).
7292 % The approach is first to tally the number of space characters then to
7293 % tally the number of non-space characters. While this is rather
7294 % sloppy I haven't found a better way to achieve the same effect,
7295 % especially given that all of the characters in |#1| have already been
7296 % assigned their category codes.
7297 % \begin{macrocode}
7298 \newcommand*{\hyxmp@string@len}[1]{%
7299 \@hyxmp@count=0
7300 \expandafter\hyxmp@count@spaces#1 {} %
7301 \expandafter\hyxmp@count@non@spaces#1{}%
7302 }
7303 % \end{macrocode}
7304 % \end{macro}
7305 %
7306 % \begin{macro}{\hyxmp@count@spaces}
7307 % Count the number of spaces in a given string. We rely on the built-in
7308 % pattern matching of \tex's |\def| primitive to pry one word at a time
7309 % off the head of the input string.
7310 % \begin{macrocode}
7311 \def\hyxmp@count@spaces#1 {%
7312 \def\hyxmp@one@token{#1}%
7313 \ifx\hyxmp@one@token\@empty
7314 \advance\@hyxmp@count by -1
7315 \else
7316 \advance\@hyxmp@count by 1
7317 \expandafter\hyxmp@count@spaces
7318 \fi
7319 }
7320 % \end{macrocode}
7321 % \end{macro}
7322 %
7323 % \begin{macro}{\hyxmp@count@non@spaces}
7324 % Count the number of non-spaces in a given string. Ideally, we'd count
7325 % both spaces and non-spaces but \tex\ won't bind |#1| to a space
7326 % character (category code~10). Hence, in each iteration, |#1| is bound
7327 % to the next non-space character only.
7328 % \begin{macrocode}
7329 \newcommand*{\hyxmp@count@non@spaces}[1]{%
7330 \def\hyxmp@one@token{#1}%
7331 \ifx\hyxmp@one@token\@empty
7332 \else
7333 \advance\@hyxmp@count by 1
7334 \expandafter\hyxmp@count@non@spaces
7335 \fi
7336 }
7337 % \end{macrocode}
7338 % \end{macro}
7339 %
7340 % \subsubsection{Embedding using \XeTeX}
7341 %
7342 % \begin{macro}{\hyxmp@embed@packet@xetex}
7343 % Embed the \acro{XMP} packet using \cmdname{xdvipdfmx}-specific
7344 % |\special| commands. I don't know how to tell \cmdname{xdvipdfmx}
7345 % always to leave the \pdfterm{Metadata} stream uncompressed, so the
7346 % \acro{XMP} metadata is likely to be missed by non-\acro{PDF}-aware
7347 % \acro{XMP} viewers.
7348 % \begin{macrocode}
7349 \newcommand*{\hyxmp@embed@packet@xetex}{%
7350 \special{pdf:stream @hyxmp@Metadata (\hyxmp@xml)
7351 <<
7352 /Type /Metadata
7353 /Subtype /XML
7354 >>
7355 }%
7356 \special{pdf:put @catalog
7357 <<
7358 /Metadata @hyxmp@Metadata
7359 >>
7360 }%
7361 }
7362 % \end{macrocode}
7363 % \end{macro}
7364 %
7365 %
7366 % \subsection{Final clean-up}
7367 % \label{sec:clean-up}
7368 %
7369 % As explained in Section~\ref{sec:initial-prep}, all invocations of
7370 % \cs{AtEndPreamble} have been stored in \cs{hyxmp@aep@toks} rather than
7371 % executed. Now that \pkgname{hyperxmp} has been initialized
7372 % completely, it is finally safe to execute the accumulated
7373 % \cs{AtEndPreamble} stanzas.
7374 % \begin{macrocode}
7375 \the\hyxmp@aep@toks
7376 % \end{macrocode}
7377 %
7378 % Having saved the category code of
7379 % ``\,{\fontencoding{T1}\selectfont\textquotedbl}\,'' at the start of
7380 % the package code (Section~\ref{sec:initial-prep}), we now restore that
7381 % character's original category code.
7382 %
7383 % \begin{macrocode}
7384 \catcode`\"=\hyxmp@dq@code
7385 % \end{macrocode}
7386 %
7387 % \iffalse
7388 %</package>
7389 % \fi
7390 %
7391 % \iffalse
7392 %<*doc-helper>
7393 % \fi
7394 %
7395 % ^^A There's not much point including the documentation helper code
7396 % ^^A in the documentation itself, but feel free to remove the
7397 % ^^A following |\iffalse| and matching |\fi| if you'd like to include
7398 % ^^A it.
7399 % \iffalse
7400 %
7401 % \section{Documentation helper code}
7402 %
7403 % The following code is used to make |hyperxmp.pdf| compliant with the
7404 % \PDFstd{A}{1}{b}{} standard, at least when built from \pdfLaTeX\ or
7405 % \LuaLaTeX, but not currently \XeLaTeX\@. We include this code in a
7406 % separate file because (a)~it is less distracting squirreled away here
7407 % than placed at the top of the |.dtx| file, and (b)~putting it in a
7408 % separate file makes it easy to disable. One simply needs to comment
7409 % out the |\input{hyperxmp-stds}| line to produce an ordinary \acro{PDF}
7410 % file, which offers the benefit of live hyperlinks in almost all
7411 % \acro{PDF} readers.
7412 %
7413 %
7414 % \subsection{Ensuring proper support}
7415 %
7416 % \begin{macro}{\next}
7417 % We can't proceed without a color profile. For this document we use a
7418 % \acro{CMYK} profile, specifically uncoated FOGRA29L, as a small and
7419 % thus safe color space. As long as |FOGRA29L_uncoated.icc| lies in the
7420 % \tex\ search path (which normally includes the current directory),
7421 % we'll be able to find it.
7422 % \begin{macrocode}
7423 \let\next=\endinput
7424 \IfFileExists{FOGRA29L_uncoated.icc}
7425 {\let\next=\relax}
7426 {}
7427 \next
7428 % \end{macrocode}
7429 % \end{macro}
7430 %
7431 % \begin{macro}{\next}
7432 % We can't proceed unless we're running either \pdfLaTeX\ or
7433 % \LuaLaTeX\@. \XeLaTeX\ doesn't appear to support all of the
7434 % mechanisms required below to support \PDFstd{A}{1}{b}{} compliance.
7435 % If you know how to adapt the following code to \XeLaTeX, please
7436 % contact the \pkgname{hyperxmp} maintainer.
7437 % \begin{macrocode}
7438 \let\next=\endinput
7439 \ifPDFTeX
7440 \let\next=\relax
7441 \else
7442 \ifLuaTeX
7443 \usepackage{luatex85}
7444 \let\next=\relax
7445 \fi
7446 \fi
7447 \next
7448 % \end{macrocode}
7449 % \end{macro}
7450 %
7451 % If we're here, we should be able to make |hyperxmp.pdf|
7452 % \PDFstd{A}{1}{b}{}-compliant.
7453 % \begin{macrocode}
7454 \def\wantpdfstandards{}
7455 \typeout{Generating PDF/A-1b compliant hyperxmp documentation.}
7456 % \end{macrocode}
7457 %
7458 %
7459 % \subsection{Complying with the PDF/A-1b standard}
7460 %
7461 % We can't specify \PDFstd{A}{1}{b}{} compliance until after both
7462 % \pkgname{hyperref} and \pkgname{hyperxmp} are loaded so we do so at
7463 % the beginning of the document.
7464 % \begin{macrocode}
7465 \AtBeginDocument{%
7466 \hypersetup{%
7467 pdfapart=1,
7468 pdfaconformance=b
7469 }%
7470 }
7471 % \end{macrocode}
7472 %
7473 % \PDFstd{A}{1}{b}{} forbids object compression. This is normally the
7474 % default for \pdfLaTeX, but the \TeX\ Live distribution changes the
7475 % default to compression level~2. Here we explicitly change it back
7476 % to~0.
7477 % \begin{macrocode}
7478 \pdfobjcompresslevel=0
7479 % \end{macrocode}
7480 %
7481 % We specify an uncoated FOGRA29 output intent for both
7482 % \PDFstd{A}{1}{b}{} and \acro{PDF/X} (all versions).
7483 % \begin{macrocode}
7484 \immediate\pdfobj stream attr {/N 4} file {FOGRA29L_uncoated.icc}
7485 \edef\iccobj{\the\pdflastobj}
7486 \pdfcatalog{%
7487 /OutputIntents [
7488 <<
7489 /Type /OutputIntent
7490 /S /GTS_PDFA1
7491 /DestOutputProfile \iccobj\space 0 R
7492 /OutputConditionIdentifier (Uncoated FOGRA29)
7493 /Info (FOGRA29L)
7494 >>
7495 <<
7496 /Type /OutputIntent
7497 /S /GTS_PDFX
7498 /DestOutputProfile \iccobj\space 0 R
7499 /OutputConditionIdentifier (Uncoated FOGRA29)
7500 /Info (FOGRA29L)
7501 >>
7502 ]
7503 }
7504 % \end{macrocode}
7505 %
7506 % \begin{macro}{\HyColor@HyperrefBorderColor}
7507 % We just specified a \acro{CMYK} color profile, but \pkgname{hyperref}
7508 % normally produces \acro{RGB} link borders. The following code
7509 % instructs \pkgname{hyperref} to accept \acro{CMYK} colors.
7510 % \begin{macrocode}
7511 \makeatletter
7512 \AtBeginDocument{%
7513 \let\HyColor@HyperrefBorderColor\HyColor@XZeroOneThreeFour
7514 }
7515 \makeatother
7516 % \end{macrocode}
7517 % \end{macro}
7518 %
7519 % \pkgname{hyperref}'s default colors are specified as \acro{RGB}\@. We
7520 % need to reassign them so they take on \acro{CMYK} values. (Note that
7521 % the main document already loaded the \pkgname{xcolor} package with the
7522 % \optname{cmyk} option.) Annoyingly, very few \acro{PDF} readers
7523 % correctly color hyperlinks defined with \acro{CMYK} link borders. At
7524 % the time of this writing (March 2020), Adobe Acrobat DC displays the
7525 % \emph{reverse} of the specified color, presumably because it always
7526 % interprets the first three color channels as red, green, and blue.
7527 % Other browsers display no border or a black border. This is not a
7528 % huge problem in practice because (a)~it's more important to be able to
7529 % distinguish hyperlinks from regular text than to render with correct
7530 % border colors and (b)~\acro{PDF} readers may suppress hyperlink
7531 % borders anyway for \PDFstd{A}{1}{b}{} documents.
7532 % \begin{macrocode}
7533 \AtBeginDocument{%
7534 \hypersetup{%
7535 citebordercolor=green,
7536 linkbordercolor=red,
7537 urlbordercolor=cyan
7538 }%
7539 }
7540 % \end{macrocode}
7541 %
7542 % \begin{macro}{\@tempdima}
7543 % \begin{macro}{\boxwd}
7544 % \begin{macro}{\boxht}
7545 % \begin{macro}{\next}
7546 % Explicitly specify the \pdfterm{BleedBox} and \pdfterm{TrimBox} to use
7547 % for every page of the documentation.
7548 % \begin{macrocode}
7549 \makeatletter
7550 \@tempdima=0.996264\paperwidth
7551 \edef\boxwd{\strip@pt\@tempdima}
7552 \@tempdima=0.996264\paperheight
7553 \edef\boxht{\strip@pt\@tempdima}
7554 \makeatother
7555 \edef\next{%
7556 \protect\pdfpageattr{
7557 /BleedBox [0.0 0.0 \boxwd\space \boxht]
7558 /TrimBox [0.0 0.0 \boxwd\space \boxht]
7559 /StructParents 0
7560 /Tabs /S
7561 }%
7562 }
7563 \next
7564 % \end{macrocode}
7565 % \end{macro}
7566 % \end{macro}
7567 % \end{macro}
7568 % \end{macro}
7569 %
7570 % \fi
7571 %
7572 % \iffalse
7573 %</doc-helper>
7574 % \fi
7575 %
7576 % \Finale
7577 \endinput