From 9f10616edfed3b801246055ec6d06631ff60f9b1 Mon Sep 17 00:00:00 2001 From: Norbert Preining Date: Sun, 21 Jun 2020 03:02:27 +0000 Subject: CTAN sync 202006210302 --- macros/latex/contrib/hyperxmp/hyperxmp.dtx | 491 ++++++++++++++++++++--------- macros/latex/contrib/hyperxmp/hyperxmp.pdf | Bin 1326332 -> 1341156 bytes 2 files changed, 338 insertions(+), 153 deletions(-) (limited to 'macros/latex/contrib/hyperxmp') diff --git a/macros/latex/contrib/hyperxmp/hyperxmp.dtx b/macros/latex/contrib/hyperxmp/hyperxmp.dtx index e4e4f52837..9861784479 100644 --- a/macros/latex/contrib/hyperxmp/hyperxmp.dtx +++ b/macros/latex/contrib/hyperxmp/hyperxmp.dtx @@ -22,7 +22,7 @@ %\NeedsTeXFormat{LaTeX2e}[1999/12/01] %\ProvidesPackage{hyperxmp} %<*package> - [2020/06/13 v5.3 Store hyperref metadata in XMP format] + [2020/06/19 v5.4 Store hyperref metadata in XMP format] % % %<*driver> @@ -112,7 +112,7 @@ % % \fi % -% \CheckSum{2610} +% \CheckSum{2671} % % \CharacterTable % {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z @@ -190,8 +190,16 @@ % % ^^A Define a few logical styles. % \DeclareRobustCommand{\term}[1]{#1\SortIndex{#1}{#1}} -% \DeclareRobustCommand{\pkgname}[1]{\mbox{\textsf{#1}}\SortIndex{#1}{\textsf{#1}}} -% \DeclareRobustCommand{\clsname}[1]{\mbox{\textsf{#1}}\SortIndex{#1}{\textsf{#1}}} +% \DeclareRobustCommand{\pkgname}[1]{^^A +% \mbox{\textsf{#1}}^^A +% \SortIndex{#1}{\textsf{#1} (package)}^^A +% \index{packages=\textsf{#1}}^^A +% } +% \DeclareRobustCommand{\clsname}[1]{^^A +% \mbox{\textsf{#1}}^^A +% \SortIndex{#1}{\textsf{#1} (class)}^^A +% \index{classes=\textsf{#1}}^^A +% } % \makeatletter % \DeclareRobustCommand{\xmpprop}[2][]{^^A XMP property % \def\xmppropopt{#1}^^A @@ -210,7 +218,11 @@ % } % \makeatother % \index{XMP=\acrostyle {XMP}&properties|see{properties, \acrostyle{XMP}}} -% \DeclareRobustCommand{\pdfterm}[1]{\mbox{\textsf{#1}}\SortIndex{#1}{\textsf{#1}}} +% \DeclareRobustCommand{\pdfterm}[1]{^^A +% \mbox{\textsf{#1}}^^A +% \SortIndex{#1}{\textsf{#1}}^^A +% \index{PDF=\acrostyle{PDF}=\textsf{#1}}^^A +% } % \DeclareRobustCommand{\cmdname}[1]{\mbox{\texttt{#1}}\SortIndex{#1}{\texttt{#1}}} % \DeclareRobustCommand{\optname}[1]{^^A % \mbox{\textsf{#1}}^^A @@ -272,6 +284,15 @@ % \DeclareRobustCommand{\Lua}{^^A % Lua\index{Lua}^^A % } +% \DeclareRobustCommand{\Koma}{^^A +% Koma^^A +% \SortIndex{Koma}{Koma (class)}^^A +% \index{classes&Koma}^^A +% } +% \DeclareRobustCommand{\ACMclass}{^^A +% \acro{ACM}^^A +% \index{classes&ACM=\acrostyle{ACM}}^^A +% } % % ^^A Define an environment just like macro but for Lua functions. % \makeatletter @@ -1143,15 +1164,16 @@ % \optname{pdfmetalang} defaults to the same value as \optname{pdflang} % if non-empty, ``|x-default|'' otherwise. \pkgname{hyperxmp} % recognizes some class-specific metadata as well, such as that provided -% via the Koma letter classes (e.g.,~\clsname{scrlttr2}) and the -% \acro{ACM} article class (\clsname{acmart}). +% via the \Koma\ letter classes (e.g.,~\clsname{scrlttr2}) and the +% \ACMclass\ article class (\clsname{acmart}). % -% If a document uses the \pkgname{polyglossia} package, it is -% recommended that it \emph{not} explicitly set \optname{pdflang}. -% \optname{pdflang} accepts only a single language name while -% \pkgname{hyperxmp} can automatically query \pkgname{polyglossia} for a -% list of all languages used in the document and include this list in an -% \acro{XMP} \xmpterm{dc:language} element. +% If a document uses either or \pkgname{babel} or \pkgname{polyglossia} +% package, it is recommended that it \emph{not} explicitly set +% \optname{pdflang}. \optname{pdflang} accepts only a single language +% name while \pkgname{hyperxmp} can automatically query \pkgname{babel} +% and \pkgname{polyglossia} for a list of all languages used in the +% document and include this list in an \acro{XMP} \xmpprop{dc:language} +% element. % % \usagenote{Multilingual metadata} % \label{note:multilingual} @@ -2212,6 +2234,7 @@ }{}% }% % \end{macrocode} +% \Needspace{6\baselineskip} % \begin{macro}{\KV@Hyp@pdfauthor} % \begin{macro}{\xmpcomma} % \begin{macro}{\xmpquote} @@ -2337,6 +2360,26 @@ % \end{macro} % \end{macro} % +% \begin{macro}{\hyxmp@concated@metadata} +% Assume that if the document loaded either \pkgname{babel} or +% \pkgname{polyglossia} it will eventually define one or more languages +% that \pkgname{hyperxmp} can list within a \xmpprop{dc:language} +% element. +% \begin{macrocode} +\edef\hyxmp@concated@metadata{} +\AtEndPreamble{% + \@ifpackageloaded{babel}{% + \edef\hyxmp@concated@metadata{babel}% + }{% + \@ifpackageloaded{polyglossia}{% + \edef\hyxmp@concated@metadata{polyglossia}% + }{% + }% + }% +} +% \end{macrocode} +% \end{macro} +% % \begin{macro}{\hyxmp@find@metadata} % \begin{macro}{\hyxmp@concated@metadata} % Issue a warning message if the author failed to specify any metadata @@ -2348,6 +2391,7 @@ % \begin{macrocode} \newcommand*{\hyxmp@find@metadata}{% \edef\hyxmp@concated@metadata{% + \hyxmp@concated@metadata \@baseurl \@pdfauthor \@pdfauthortitle @@ -2452,11 +2496,151 @@ % \end{macro} % \end{macro} % +% Rather than load \pkgname{hyperref} ourself we let the author do it +% then verify he actually did. This approach gives the author the +% flexibility to load \pkgname{hyperxmp} and \pkgname{hyperref} in +% either order and to call |\hypersetup| anywhere in the document's +% preamble, not just before \pkgname{hyperxmp} is loaded. +% \begin{macrocode} +\AtEndPreamble{% + \@ifpackageloaded{hyperref}{% +% \end{macrocode} +% If \cs{@pdflang} is not set, see if we can detect the document +% language via either the \pkgname{babel} or \pkgname{polyglossia} +% packages. +% \begin{macrocode} + \@if@def@and@nonempty{@pdflang}{% + }{% + \hyxmp@detect@langs + }% +% \end{macrocode} +% Fill in any missing metadata we can using values provided by the author +% via mechanisms other than the \cs{hypersetup} command. +% \begin{macrocode} + \hyxmp@auto@assign@data +% \end{macrocode} +% Most \acro{PDF} standards dictate that if the same metadata appear in +% both the \acro{XMP} packet and the \acro{PDF} \pdfterm{Info} +% dictionary, the metadata must match. This requirement poses a problem +% for a user-unspecified \optname{pdfcreationdate} in the context of +% \XeLaTeX\@. In this case we explicitly define |\@pdfcreationdate| as +% |\hyxmp@today@pdf| to prevent the \cmdname{xdvipdfmx} back-end +% processor from detecting a missing \pdfterm{CreationDate} in the +% \pdfterm{Info} dictionary and adding its own---typically a few seconds +% after \pkgname{hyperxmp} has constructed an \xmpprop{xmp:CreateDate} +% for the \acro{XMP} metadata and leading to a metadata mismatch. +% \begin{macrocode} + \@ifundefined{XeTeXversion}{}{% + \@ifmtargexp{\@pdfcreationdate}{% + \let\@pdfcreationdate=\hyxmp@today@pdf + }% + {}% + }% +% \end{macrocode} +% If the document claims to comply with one or more \acro{PDF} +% standards, check that all of the requisite metadata are present. +% \begin{macrocode} + \hyxmp@check@standards +% \end{macrocode} +% Older versions of \pkgname{hyperref} write the \pdfterm{Info} +% dictionary to the \acro{PDF} file at the end of the document. New +% versions of \pkgname{hyperref} write the \pdfterm{Info} dictionary to +% the \acro{PDF} file at the \emph{beginning} of the document. For +% compatibility with both old and new \pkgname{hyperref} implementations +% we suppress writing the \pdfterm{Info} dictionary here, at the +% beginning of the document. +% \changes{v4.1}{2019/04/02}{Invoke +% \protect\cs{hyxmp@no@info@lists} at the beginning of the +% document, for compatibility with both newer and older versions of +% \protect\pkgname{hyperref}} +% \begin{macrocode} + \hyxmp@no@info@lists +% \end{macrocode} +% We wait until the end of the document to construct the \acro{XMP} +% packet and write it to the \acro{PDF} document catalog. This gives +% the author ample opportunity to provide metadata to \pkgname{hyperref} +% and thereby \pkgname{hyperxmp}. +% \begin{macrocode} + \hyxmp@at@end{% + \hyxmp@find@metadata + \hyxmp@embed@packet + }% + }{% + \PackageWarningNoLine{hyperxmp}{% + \jobname.tex failed to include a\MessageBreak + \string\usepackage\string{hyperref\string} + in the preamble.\MessageBreak + Consequently, all hyperxmp functionality will be\MessageBreak + disabled}% + }% +} +% \end{macrocode} +% +% A number of classes either require or recommend that authors declare +% various class-specific metadata \emph{after} the |\begin{document}|. +% So where do we invoke \cs{hyxmp@auto@assign@data}? On one hand, we +% want to invoke it before the |\begin{document}| because this may +% obviate \pkgname{hyperxmp}'s ``\meta{job name} did not specify any +% metadata'' warning message. On the other hand, we want to invoke it +% after the |\begin{document}| so it picks up metadata not specified in +% the preamble. Our solution is to invoke \cs{hyxmp@auto@assign@data} +% twice: both before the |\begin{document}| and at the |\end{document}|. +% We additionally identify here the natural language(s) in use because +% these aren't known until the end of the document when provided by +% \pkgname{babel} or \pkgname{polyglossia}. Also, \pkgname{hyperref} +% forbids changes to \optname{pdflang} past the |\begin{document}|. +% \changes{v1.3}{2011/04/25}{Introduced the \protect\optname{pdfmetalang} +% package option, which enables an author to specify the language in which he +% wrote the document's metadata} +% \changes{v2.0}{2012/08/02}{New \cs{AtBeginDocument} code from Heiko +% Oberdiek to properly encode \cs{@pdfmetalang}} +% \changes{v3.3}{2017/07/21}{Don't overwrite an existing +% \protect\optname{pdfmetalang} with \protect\optname{pdflang} or +% \protect\texttt{x-default}. This addresses a bug report by Niklas +% Beisert} +% \changes{v5.4}{2020/06/18}{Moved the automatic assignment of +% \protect\cs{@pdflang} and \protect\cs{@pdfmetalang} from +% \protect\cs{hyxmp@auto@assign@data} to within a call to +% \protect\cs{hyxmp@at@end}} +% \begin{macrocode} +\hyxmp@at@end{% + \hyxmp@set@dc@lang + \ifx\@pdfmetalang\@empty + \ifx\@pdflang\@empty + \let\@pdfmetalang=\hyxmp@x@default + \else + \edef\@pdfmetalang{\@pdflang}% + \fi + \fi + \hyxmp@xmlify\@pdfmetalang + \hyxmp@auto@assign@data +} +% \end{macrocode} +% +% +% \subsection{Advanced metadata detection} +% +% \pkgname{hyperxmp} strives to be as convenient and user-friendly as +% possible. To that end, we try to automatically detect as much +% metadata as possible. The author can of course augment or override +% autodetected metadata by explicitly providing values to +% \cs{hypersetup}, but the hope is that we can save the author some +% effort in many cases. +% +% In this section, we identify additional metadata we can use. Most of +% the functionality is class- or package-specific. For example, we +% check for phone numbers provided to the \Koma\ letter classes via +% |\setkomavar{fromphone}{|\dots|}| and/or +% |\setkomavar{frommobilephone}{|\dots|}|, street addresses provided to +% the \ACMclass\ article class via \cs{affiliation}, and languages +% the \pkgname{polyglossia} package is instructed to load via +% \cs{setdefaultlanguage} and \cs{setotherlanguage}. +% % \begin{macro}{\hyxmp@set@koma@phones} % \changes{v5.3}{2020/06/07}{Added this macro} % \begin{macro}{\hyxmp@koma@phones} % Define \cs{hyxmp@koma@phones} as a comma-separated list of the phone -% numbers provided to a Koma letter class (mobile and landline). +% numbers provided to a \Koma\ letter class (mobile and landline). % \begin{macrocode} \newcommand*{\hyxmp@set@koma@phones}{% \@if@def@and@nonempty{scr@frommobilephone@var}{% @@ -2525,30 +2709,6 @@ % \begin{macrocode} \newcommand*{\hyxmp@auto@assign@data}{% % \end{macrocode} -% If the author explicitly specified the language to use for the -% document's metadata, we use that. If not, we use the document -% language, specified to \pkgname{hyperref} with the \protect\optname{pdflang} -% option. If the author did not specify a language, we use |x-default| -% as the metadata language. -% \changes{v1.3}{2011/04/25}{Introduced the \protect\optname{pdfmetalang} -% package option, which enables an author to specify the language in which he -% wrote the document's metadata} -% \changes{v2.0}{2012/08/02}{New \cs{AtBeginDocument} code from Heiko -% Oberdiek to properly encode \cs{@pdfmetalang}} -% \changes{v3.3}{2017/07/21}{Don't overwrite an existing -% \protect\optname{pdfmetalang} with \protect\optname{pdflang} or -% \protect\texttt{x-default}. This addresses a bug report by Niklas -% Beisert} -% \begin{macrocode} - \ifx\@pdfmetalang\@empty - \ifx\@pdflang\@empty - \let\@pdfmetalang=\hyxmp@x@default - \else - \edef\@pdfmetalang{\@pdflang}% - \fi - \fi - \hyxmp@xmlify\@pdfmetalang -% \end{macrocode} % If the author left \optname{pdftitle} blank but specified |\title|, % use the title for \optname{pdftitle}. Do likewise for various other % metadata: identify author-provided information that can be co-opted @@ -2558,7 +2718,7 @@ % unspecified. Thanks to Maciej Radziejewski for the suggestion} % \changes{v5.3}{2020/05/29}{Consider other author-provided sources of % metadata. Thanks to Robin Schwab for proposing that -% \protect\pkgname{hyperxmp} use the Koma letter classes's metadata} +% \protect\pkgname{hyperxmp} use the \Koma\ letter classes's metadata} % \begin{macrocode} \hyxmp@use@first@valid{pdftitle}{\@pdftitle}{% \scr@subject@var,% @@ -2728,6 +2888,8 @@ % ``10.1145/nnnnnnn.nnnnnnn'', or the example \acro{DOI} specified in % the \clsname{acmart} example document, ``10.1145/1122445.1122456''. % We ignore both of those \acro{DOI}s. +% \changes{v5.4}{2020/06/16}{Bug fix: Correct a missing ``else'' argument +% in two invocations of \protect\cs{@if@def@and@nonempty}} % \begin{macrocode} \@if@def@and@nonempty{@acmDOI}{% \IfSubStr{\@acmDOI}{10.1145/1122445.1122456}{}{% @@ -2738,6 +2900,7 @@ }% }% }% + {}% % \end{macrocode} % \begin{macro}{\hyxmp@strip@isbn@date} % \begin{macro}{\hyxmp@acm@isbn} @@ -2761,6 +2924,7 @@ }% }% }% + {}% % \end{macrocode} % \end{macro} % \end{macro} @@ -2819,16 +2983,17 @@ % \begin{macro}{\hyxmp@dc@lang} % \cs{@pdflang} is used in both the \acro{PDF} document catalog (the % \pdfterm{Lang} key, written to the \acro{PDF} file by -% \pkgname{hyperref}) and in the Dublin Core \xmpterm{dc:language} tag +% \pkgname{hyperref}) and in the Dublin Core \xmpprop{dc:language} tag % (Section~\ref{sec:dublin-core}). Normally, these are the same. % However, \pdfterm{Lang} accepts only a single language while -% \xmpterm{dc:language} accepts multiple languages. If the document -% loads \pkgname{polyglossia} and does not specify \optname{pdflang} in -% \cs{hypersetup}, \cs{hyxmp@set@dc@lang} is redefined below to set -% \cs{hyxmp@dc@lang} to \pkgname{polyglossia}'s list of used languages -% (once this is known). Otherwise, \cs{hyxmp@set@dc@lang} assigns the -% language specified by \optname{pdflang} (if any) to -% \cs{hyxmp@dc@lang}. +% \xmpprop{dc:language} accepts multiple languages. If the document +% loads either \pkgname{babel} or \pkgname{polyglossia} and does not +% specify \optname{pdflang} in \cs{hypersetup}, \cs{hyxmp@set@dc@lang} +% is redefined below (in \cs{hyxmp@detect@langs}) to set +% \cs{hyxmp@dc@lang} to \pkgname{babel}\slash\pkgname{polyglossia}'s +% list of used languages (once this is known). Otherwise, +% \cs{hyxmp@set@dc@lang} assigns the language specified by +% \optname{pdflang} (if any) to \cs{hyxmp@dc@lang}. % \begin{macrocode} \newcommand*{\hyxmp@set@dc@lang}{% \let\hyxmp@dc@lang=\@pdflang @@ -2837,119 +3002,139 @@ % \end{macro} % \end{macro} % -% Rather than load \pkgname{hyperref} ourself we let the author do it -% then verify he actually did. This approach gives the author the -% flexibility to load \pkgname{hyperxmp} and \pkgname{hyperref} in -% either order and to call |\hypersetup| anywhere in the document's -% preamble, not just before \pkgname{hyperxmp} is loaded. +% \begin{macro}{\hyxmp@detect@langs} +% If \optname{pdflang} was not specified, try to determine the document +% language(s) using either \pkgname{babel}'s or \pkgname{polyglossia}'s +% definitions. If so, we redefine \cs{hyxmp@set@dc@lang} as a +% comma-separate list of languages used in the document. +% \changes{v5.3}{2020/06/08}{Acquire the default language from the +% \protect\pkgname{polyglossia} package, if loaded. Thanks to +% Robin Schwab for bringing that package to my attention} +% \changes{v5.4}{2020/06/16}{Refactored language detection into a +% separate command} % \begin{macrocode} -\AtEndPreamble{% - \@ifpackageloaded{hyperref}{% +\newcommand*{\hyxmp@detect@langs}{% + \@ifundefined{mainbcp47id}{% + \@ifundefined{LocaleForEach}{% % \end{macrocode} -% In older versions of \pkgname{hyperref}, \cs{@pdflang} is set to +% The document doesn't appear to have loaded either \pkgname{babel} or +% \pkgname{polyglossia}. In this case we have one small task to do. In +% older versions of \pkgname{hyperref}, \cs{@pdflang} is set to % \cs{@empty} if \optname{pdflang} is not specified. In newer versions % of \pkgname{hyperref}, \cs{@pdflang} is set to \cs{relax} if % \optname{pdflang} is not specified. The latter is a bit problematic % for \pkgname{hyperxmp} because it makes \cs{@pdflang} non-expandable, % which causes a literal ``\cs{@pdflang}'' to be written as \acro{XMP} -% metadata. To avoid that situation we redefine \cs{@pdflang} as -% \cs{@empty} if we see it set to \cs{relax}. Actually, we first check -% if the \pkgname{polyglossia} package has specified the document's main -% language in \cs{mainbcp47id}. If so, we use that instead of -% \cs{@empty} and redefine \cs{hyxmp@set@dc@lang} to consider -% \pkgname{polyglossia}'s list of used languages. +% metadata. To avoid that situation we explicitly set \cs{@pdflang} to +% \cs{@empty} to avoid problems with non-expandable symbols. % \changes{v2.3a}{2013/04/16}{Bug fix: Redefine \cs{@pdflang} as % \cs{@empty} when \protect\pkgname{hyperref} has set % it to \cs{relax}} -% \changes{v5.3}{2020/06/08}{Acquire the default language from the -% \protect\pkgname{polyglossia} package, if loaded. Thanks to -% Robin Schwab for bringing that package to my attention} % \begin{macrocode} - \ifx\@pdflang\relax - \@ifundefined{mainbcp47id}{% - \let\@pdflang=\@empty - }{% - \xdef\@pdflang{\csname mainbcp47id\endcsname}% - \renewcommand*{\hyxmp@set@dc@lang}{% - \edef\hyxmp@dc@lang{\xpg@bcp@loaded}% + \let\@pdflang=\@empty + }{% +% \end{macrocode} +% Use \pkgname{babel}'s \cs{LocaleForEach} and \cs{getlocaleproperty} to +% construct a comma-separate list of all document languages. We also +% set \cs{@pdflang} within \cs{hyxmp@set@dc@lang}. +% (\cs{hyxmp@detect@langs} should only be called if \cs{@pdflang} is +% empty.) Unlike \pkgname{polyglossia}, \pkgname{babel} does not +% provide information about the main language until after the +% |\begin{document}| so we have to wait until \cs{hyxmp@set@dc@lang} is +% called to set it. +% \changes{v5.4}{2020/06/17}{Added support for \protect\pkgname{babel}} +% \begin{macrocode} + \BabelEnsureInfo + \renewcommand*{\hyxmp@set@dc@lang}{% + \getlocaleproperty\@pdflang{\bbl@main@language}{identification/tag.bcp47}% + \hyxmp@write@pdflang + \let\hyxmp@dc@lang=\relax + \LocaleForEach{% + \getlocaleproperty\hyxmp@lang@tag{####1}{identification/tag.bcp47}% + \ifx\hyxmp@dc@lang\relax + \edef\hyxmp@dc@lang{\hyxmp@lang@tag}% + \else + \edef\hyxmp@dc@lang{\hyxmp@dc@lang,\hyxmp@lang@tag}% + \fi }% }% - \fi -% \end{macrocode} -% Fill in any missing metadata we can using values provided by the author -% via mechanisms other than the \cs{hypersetup} command. -% \begin{macrocode} - \hyxmp@auto@assign@data + }% + }{% % \end{macrocode} -% Most \acro{PDF} standards dictate that if the same metadata appear in -% both the \acro{XMP} packet and the \acro{PDF} \pdfterm{Info} -% dictionary, the metadata must match. This requirement poses a problem -% for a user-unspecified \optname{pdfcreationdate} in the context of -% \XeLaTeX\@. In this case we explicitly define |\@pdfcreationdate| as -% |\hyxmp@today@pdf| to prevent the \cmdname{xdvipdfmx} back-end -% processor from detecting a missing \pdfterm{CreationDate} in the -% \pdfterm{Info} dictionary and adding its own---typically a few seconds -% after \pkgname{hyperxmp} has constructed an \xmpprop{xmp:CreateDate} -% for the \acro{XMP} metadata and leading to a metadata mismatch. +% Use \pkgname{polyglossia}'s \cs{mainbcp47id} as the document's main +% language and its \cs{xpg@bcp@loaded} as a comma-separated list of all +% document languages. % \begin{macrocode} - \@ifundefined{XeTeXversion}{}{% - \@ifmtargexp{\@pdfcreationdate}{% - \let\@pdfcreationdate=\hyxmp@today@pdf - }% - {}% + \xdef\@pdflang{\csname mainbcp47id\endcsname}% + \renewcommand*{\hyxmp@set@dc@lang}{% + \edef\hyxmp@dc@lang{\xpg@bcp@loaded}% }% + }% +} % \end{macrocode} -% If the document claims to comply with one or more \acro{PDF} -% standards, check that all of the requisite metadata are present. +% \end{macro} +% +% \begin{macro}{\hyxmp@write@pdflang} +% Here's a conundrum: \pkgname{babel} doesn't provide access to +% information on the document's main language until after the +% |\begin{document}|. However, \pkgname{hyperref} allows +% \optname{pdflang} to be set only \emph{before} the |\begin{document}|. +% So what do we do? +% +% Note that the underlying \cs{@pdflang} macro serves two purposes: +% \pkgname{hyperref} uses it to set the \pdfterm{Lang} entry in the +% \acro{PDF} document catalog, and \pkgname{hyperxmp} uses it as the +% default metadata language. The latter is used in \xmpprop{dc:title} +% and other \acro{XMP} properties that can include per-language +% variants. We set \cs{@pdflang} once we have the main-language +% information from \pkgname{babel}. As this is too late for +% \pkgname{hyperref}, we bypass \pkgname{hyperref} and manually write +% the \pdfterm{Lang} key into the document catalog. This is a bit +% kludgy, but we do leverage as much of \pkgname{hyperref}'s +% infrastructure as possible to increase \cs{hyxmp@write@pdflang}'s +% robustness. +% \changes{v5.4}{2020/06/19}{Added this macro} +% \begin{macrocode} +\newcommand*{\hyxmp@write@pdflang}{% + \@ifundefined{pdfmark}{% + \@ifundefined{@pdfm@mark}{% + \@ifundefined{pdfcatalog}{% +% \end{macrocode} +% We don't recognize the \tex\ engine or \acro{PDF} generator. (This is +% unexpected.) We therefore issue a warning message and take no further +% action. +% \begin{macrocode} + \PackageWarning{hyperxmp}{Unknown PDF generator; not setting the language (\@pdflang) in the PDF catalog}% + }{% +% \end{macrocode} +% We're running either \pdfTeX\ or \LuaTeX\@. (\pkgname{hyperref} +% defines \cs{pdfcatalog} in the latter case.) % \begin{macrocode} - \hyxmp@check@standards + \pdfcatalog{/Lang (\@pdflang)}% + }% + }{% % \end{macrocode} -% Older versions of \pkgname{hyperref} write the \pdfterm{Info} -% dictionary to the \acro{PDF} file at the end of the document. New -% versions of \pkgname{hyperref} write the \pdfterm{Info} dictionary to -% the \acro{PDF} file at the \emph{beginning} of the document. For -% compatibility with both old and new \pkgname{hyperref} implementations -% we suppress writing the \pdfterm{Info} dictionary here, at the -% beginning of the document. -% \changes{v4.1}{2019/04/02}{Invoke -% \protect\cs{hyxmp@no@info@lists} at the beginning of the -% document, for compatibility with both newer and older versions of -% \protect\pkgname{hyperref}} +% We're running either \XeTeX\ or \tex~+ +% \texttt{dvipdfm}.\SortIndex{dvipdfm}{\texttt{dvipdfm}}. % \begin{macrocode} - \hyxmp@no@info@lists + \@pdfm@mark{docview << /Lang (\@pdflang) >>}% + }% +}{% % \end{macrocode} -% We wait until the end of the document to construct the \acro{XMP} -% packet and write it to the \acro{PDF} document catalog. This gives -% the author ample opportunity to provide metadata to \pkgname{hyperref} -% and thereby \pkgname{hyperxmp}. +% We're running either \tex~+ \Dvips~+ Adobe Acrobat Distiller (or other +% converter from PostScript to \acro{PDF}). % \begin{macrocode} - \hyxmp@at@end{% - \hyxmp@find@metadata - \hyxmp@embed@packet + \pdfmark{% + pdfmark=/PUT,Raw={% + \string{Catalog\string} << + /Lang (\@pdflang) + >> + }% }% - }{% - \PackageWarningNoLine{hyperxmp}{% - \jobname.tex failed to include a\MessageBreak - \string\usepackage\string{hyperref\string} - in the preamble.\MessageBreak - Consequently, all hyperxmp functionality will be\MessageBreak - disabled}% }% } % \end{macrocode} -% -% A number of classes either require or recommend that authors declare -% various class-specific metadata \emph{after} the |\begin{document}|. -% So where do we invoke \cs{hyxmp@auto@assign@data}? On one hand, we -% want to invoke it before the |\begin{document}| because this may -% obviate \pkgname{hyperxmp}'s ``\meta{job name} did not specify any -% metadata'' warning message. On the other hand, we want to invoke it -% after the |\begin{document}| so it picks up metadata not specified in -% the preamble. Our solution is to invoke \cs{hyxmp@auto@assign@data} -% twice: both before the |\begin{document}| at at the |\end{document}|. -% \begin{macrocode} -\hyxmp@at@end{\hyxmp@auto@assign@data} -% \end{macrocode} +% \end{macro} % % % \subsection{Manipulating author-supplied data} @@ -4628,7 +4813,7 @@ _______________________________________________________________^^J% % Assign |\hyxmp@major@minor| to be the \acro{PDF} version targeted by % the running \TeX\ engine. % \changes{v5.2}{2020/05/10}{Added this macro. \string\pkgname{hyperxmp} -% now correctly specifies \string\xmpterm{pdf:PDFVersion} when +% now correctly specifies \string\xmpprop{pdf:PDFVersion} when % generating \string\acro{PDF}~2.0+. Thanks to Ulrike Fischer for % alerting me to \string\acro{PDF}~2.0's availability in the % \string\TeX\ ecosystem and informing me how to activate it} @@ -4852,11 +5037,11 @@ ______^^J% % \end{macro} % % \begin{macro}{\hyxmp@cond@dc@identifier} -% Conditionally add a \xmpterm{dc:identifier} tag. Given a prefix +% Conditionally add a \xmpprop{dc:identifier} tag. Given a prefix % string~(|#1|) and a main string~(|#2|), wrap these in a -% \xmpterm{dc:identifier} if the main string is nonempty and +% \xmpprop{dc:identifier} if the main string is nonempty and % |\hyxmp@xmlified| \emph{is} empty (implying the -% \xmpterm{dc:identifier} has not yet been written). +% \xmpprop{dc:identifier} has not yet been written). % \changes{v5.2}{2020/04/29}{Added this macro} % \begin{macrocode} \newcommand*{\hyxmp@cond@dc@identifier}[2]{% @@ -4883,14 +5068,19 @@ ______^^J% % the author specified \optname{pdftype}, and the % \xmpprop{dc:identifier} if the author specified % \optname{pdfidentifier} or if we can derive it from other options. We -% also specify the \xmpprop{dc:date} property using the date the -% document was run through \LaTeX\ and the \xmpprop{dc:source} property -% using the base name of the source file with |.tex| appended. +% also specify the \xmpprop{dc:source} property using the base name of +% the source file with |.tex| appended and the \xmpprop{dc:date} +% property using the date the document was run through \LaTeX---unless +% the author specified \optname{pdfdate}, in which case we use that. % \changes{v2.0}{2012/08/26}{Added support for \xmpprop{dc:language} % and \xmpprop{dc:source}} % \changes{v2.4}{2013/12/21}{Made \xmpprop{dc:language} a \xmpterm{Bag} % instead of an individual item so as to conform to the latest % \acro{XMP} specifications, a detail identified by Florian Breitwieser} +% \changes{v5.3}{2020/06/13}{Include all languages used in the document +% in \protect\xmpprop{dc:language}} +% \changes{v5.4}{2020/06/18}{Bug fix: Use \protect\cs{hyxmp@today@xmp} +% as the date only if \protect\cs{@pdfdatetime} is undefined} % \begin{macrocode} \newcommand*{\hyxmp@dc@schema}{% \hyxmp@add@simple{dc:format}{application/pdf}% @@ -4898,7 +5088,11 @@ ______^^J% \hyxmp@rdf@dc[\ifHy@pdfa]{description}{\@pdfsubject}% \hyxmp@rdf@dc{rights}{\@pdfcopyright}% \hyxmp@singleton@dc{publisher}{\@pdfpublisher}% - \hyxmp@singleton@dc[Seq]{date}{\hyxmp@today@xmp}% + \@ifmtargexp{\@pdfdatetime}{% + \hyxmp@singleton@dc[Seq]{date}{\hyxmp@today@xmp}% + }{% + \hyxmp@singleton@dc[Seq]{date}{\@pdfdatetime}% + }% \hyxmp@singleton@dc{type}{\@pdftype}% \hyxmp@list@to@xml[\ifHy@pdfa]{creator}{Seq}{\hyxmp@pdfauthor}% \hyxmp@list@to@xml{subject}{Bag}{\hyxmp@pdfkeywords}% @@ -4906,18 +5100,7 @@ ______^^J% \else \hyxmp@add@simple{dc:source}{\@pdfsource}% \fi -% \end{macrocode} -% Set \cs{hyxmp@dc@lang} to either the single language \cs{@pdflang} -% (which may be unspecified) or to a list of languages used in the -% document via \pkgname{polyglossia}. See -% Section~\ref{sec:hyperref-int} for the definition of -% \cs{hyxmp@set@dc@lang}. -% \changes{v5.3}{2020/06/13}{Include all languages used in the document -% in \protect\xmpterm{dc:language}} -% \begin{macrocode} - \hyxmp@set@dc@lang \hyxmp@list@to@xml{language}{Bag}{\hyxmp@dc@lang}% -% \end{macrocode} % If |\@pdfidentifier| is empty, try setting it to each of |\@pdfdoi|, % |\@pdfeissn|, |\@pdfissn|, and |\@pdfisbn|, in turn, with proper % syntactic adjustments. @@ -6254,6 +6437,7 @@ __^^J% % file, which offers the benefit of live hyperlinks in almost all % \acro{PDF} readers. % +% % \subsection{Ensuring proper support} % % \begin{macro}{\next} @@ -6298,6 +6482,7 @@ __^^J% \typeout{Generating PDF/A-1b compliant hyperxmp documentation.} % \end{macrocode} % +% % \subsection{Complying with the PDF/A-1b standard} % % We can't specify \PDFstd{A}{1}{b}{} compliance until after both diff --git a/macros/latex/contrib/hyperxmp/hyperxmp.pdf b/macros/latex/contrib/hyperxmp/hyperxmp.pdf index 28b08b0787..998059c14e 100644 Binary files a/macros/latex/contrib/hyperxmp/hyperxmp.pdf and b/macros/latex/contrib/hyperxmp/hyperxmp.pdf differ -- cgit v1.2.3