I have a latex file including a lot of files which are initially not present
but which are created in the course of the latexmk run.
\documentclass{article}
\usepackage{functional}
\usepackage{import}
\usepackage{graphicx}
\usepackage{xcolor}
\usepackage{transparent}
\usepackage{listings}
...
\begin{document}
% \lstinputlisting[language=tex, basicstyle=\tiny,
% breaklines, lastline=25,
% float, captionpos=b, label={lst:ptx},
% caption={The ptx-file for a fig-file}]{F4_01fig2dev.ptx}
\fileIfExistTF{F4_01fig2dev.ptx}{x\lstinputlisting{F4_01fig2dev.ptx}}{\input{F4_01fig2dev.ptx}}
%\lstinputlisting{F4_01fig2dev.ptx}
\import{}{F4_01fig2dev.ptx}
%\import{}{F4_04mp2mps.ptx}
\import{}{F4_03someGnuplot.ptx}
\includegraphics{F4_05someMetapost.mps}
%\import{}{F4_06svg2pdf.ptx}%
\def\svgwidth{0.5\textwidth}%
\import{}{F4_07someSvg.ptx}
\end{document}
Seemingly, latexmk runs lualatex until the first file to be included is not present.
It ends with a fatal error but goes on creating the first included file.
Then it runs lualatex second ending up in the second fatal error because of the second missing file. Again in some rule this is created and that way it goes on.
This causes one run more than there are files to be included. Additional runs may be needed to stabilize.
I think this is not satisfying. How to solve this? I wonder whether it is possible to proceed completely different? Or to make lualatex not stop for each missing file? So the number of lualatex runs does not depend on the number of included files.
Based on suggestions below, I modified my solution:
In front of \begin{document} I added
\usepackage{functional}
\usepackage{letltxmacro}
\LetLtxMacro\latexlstinputlisting\lstinputlisting
\renewcommand{\lstinputlisting}[2][]{\fileIfExistTF{#2}{\latexlstinputlisting[#1]{#2}}{\typeout{No file #2.}}}
\LetLtxMacro\lateximport\import
\renewcommand{\import}[2]{\fileIfExistTF{#2}{\lateximport{#1}{#2}}{\typeout{No file #1#2.}}}
\LetLtxMacro\latexincludegraphics\includegraphics
\renewcommand{\includegraphics}[2][]{\fileIfExistTF{#2}{\latexincludegraphics[#1]{#2}}{\typeout{No file #2.}}}
For this special file, the number of runs of lualatex is really reduced.
But I am not content.
Redefine seems quite dangerous.
For example, redefining \input naively
seems to affect \includegraphics.
Even worse, \input can be used with a file xxx.tex without ending.
Accordingly for \includegrapics.
With this normal usage, the current solution does not work.
Any suggestions?
Maybe my .latexmkrc file is helpful
#!/usr/bin/env perl
to create pdf via lualatex
#$pdflatex = 'lualatex -file-line-error %O %S';
PDF-generating modes are:
1: pdflatex, as specified by $pdflatex variable (still largely in use)
2: postscript conversion, as specified by the $ps2pdf variable (useless)
3: dvi conversion, as specified by the $dvipdf variable (useless)
4: lualatex, as specified by the $lualatex variable (best)
5: xelatex, as specified by the $xelatex variable (second best)
$pdf_mode = 4;# specifies creation of pdf via lualatex
If additional parameters must be passed to lualatex
#'svg' package.
It converts raw SVG files to the PDF+PDF_TEX combo using InkScape.
$lualatex = "lualatex --shell-escape";
note that -recorder is implicitly added
$latex2pdfOptions = "-interaction=nonstopmode -synctex=1 -shell-escape";
$lualatex = "lualatex $latex2pdfOptions %O %S";
#$postscript_mode = $dvi_mode = 0;
to configure bibtex
$bibtex_use=2;
this cannot be done according to the according latex maven plugin,
because the according parameter maxNumReRunsLatex may be set to -1
which signifies an infinite number of runs.
$max_repeat=30;
default are tex and eps, but could also be pdf and ptx and mps
Currently, all those files are given with explicit endings,
so no extensions to be added.
#add_input_ext('');
It is what it seems to be: clean inludes what was generated by cus
$cleanup_includes_cusdep_generated = 1;
$cleanup_includes_generated = 1;
#$makeindex = 'makeindex %O -o %D %S'; # the default
TBD: clarify: xdv and dvi seem to be internal.
maybe missing other extensions in conjunction with synctex
$clean_ext .= " %R.synctex.gz";
bbl does not work
#@generated_exts = (@generated_exts, 'lol', 'bbl', 'glo', 'ist')
#print "Hello!"
#foreach (@generated_exts) {
#print "Generated exts: $_\n";
#}
#print "clean_ext\n";
Here @generated_exts is ('aux', 'fls', 'log', # generated by latex already
'toc', 'lof', 'lot', 'out', # generated by latex conditionally
'idx', 'ind', 'blg', 'ilg', # concerning indices
# strange enough: nothing for bibtex
'xdv', 'bcf'
extensions ext to be deleted by latexmk -c
Note that the file names are %R.ext.
this may cause problems with extensions containing a dot.
Also this is not general enough
if the generated file deviates from %R by more than an extension.
In this case, use $clean_ext$
Here, $clean_ext is empty.
list of listings, whereas lof and lot are already present.
push @generated_exts, "lol";
push @generated_exts, "dvi", "xdv";
for beamer class
push @generated_exts, "nav", "snm", "vrb";
why are .ist and xdy not under generated_exts?
note that currently, either %R or what is present is the extension only!
this does not make sense very much.
$clean_ext .= " stateMachine.log"; does not work, because stateMachine.log is the extension!
should be under indexing
$clean_ext .= " %R.ist %R.xdy %R-.ind %R-.idx %R-.ilg %R-.ind";
add_cus_dep('fig', 'ptx', 0, 'fig2dev');
sub fig2dev {
$file = $_[0];
print("$file");
rdb_add_generated("$file.eps", "$file.pdf", "$file.ptx");
my $GenOptions = "";
my $PtxOptions = "";
my $PdfEpsOptions = "";
print("create from $file.fig");
rdb_add_generated("$file.ptx", "$file.pdf", "$file.eps");
#fig2dev -L pstex <fig2devGenOptions> <fig2devPdfEpsOptions> xxx.fig xxx.eps
#fig2dev -L pdftex <fig2devGenOptions> <fig2devPdfEpsOptions> xxx.fig xxx.pdf
#fig2dev -L pdftex_t <fig2devGenOptions> <fig2devPtxOptions> -p xxx xxx.fig xxx.ptx
my $ret1 = system("fig2dev -L pstex $GenOptions $PdfEpsOptions $file.fig $file.eps");
my $ret2 = system("fig2dev -L pdftex $GenOptions $PdfEpsOptions $file.fig $file.pdf");
my $ret3 = system("fig2dev -L pdftex_t $GenOptions $PtxOptions -p $file $file.fig $file.ptx");
return ($ret1 or $ret2 or $ret3);
}
add_cus_dep('gp', 'ptx', 0, 'gnuplot');
sub gnuplot {
$file = $_[0];
print("create from $file.gp");
rdb_add_generated("$file.ptx", "$file.pdf", "$file.eps");
my $nuplotOptions = "";
my $ret1 = system("gnuplot -e "set terminal cairolatex pdf $gnuplotOptions;
set output '$file.ptx';
load '$file.gp'"");
my $ret2 = system("gnuplot -e "set terminal cairolatex eps $gnuplotOptions;\
set output '$file.ptx';\
load '$file.gp'"");
return $ret;
}
metapost rule from http://tex.stackexchange.com/questions/37134
#add_cus_dep('mp', 'mps', 0, 'mpost');
add_cus_dep('mp', 'mps', 0, 'mpost');
sub mpost {
my $file = $_[0];
print("create from $file.mp");
rdb_add_generated("$file.mps", "$file.fls", "$file.log");
my ($name, $path) = fileparse($file);
pushd($path);
my $return = system("mpost -interaction=nonstopmode -recorder -s prologues=2 -s 'outputtemplate="%j.mps"' $name");
popd();
return $return;
}
add_cus_dep('svg', 'ptx', 0, 'inkscape');
sub inkscape {
my $file = $_[0];
print("create from $file.svg");
rdb_add_generated("$file.ptx", "$file.pdf");
my $ret1 = system("inkscape -D --export-filename=$file.pdf --export-latex $file.svg ");
#my $ret2 = system("inkscape -D --export-filename=$file.eps --export-latex $file.svg ");
#use File::Copy;
This works only for pdf, not for eps.
#unlink($file.pdf_tex) or die "cannot unlink $file.pdf_tex";
rename("$file.pdf_tex", "$file.ptx");# or die "cannot move $file.pdf_tex";
return $ret1;# or $ret2;
}
graphics for xfig (not appropriate for mixed tex/pdf)
add_cus_dep('fig', 'pdf', 0, 'fig2pdf');
sub fig2pdf {
system( "fig2dev -Lpdf "$[0].fig" "$[0].pdf"" );
}
use splitindex
$makeindex = 'internal splitindex';
sub splitindex {
Use splitindex instead of makeindex.
The splitindex programe starts from an .idx file, makes a set of
other .idx files for separate indexes, and then runs makeindex to
make corresponding .ind files.
However, it is possible that the document uses the splitindex
package, but in a way compatible with the standard methods
compatible with makeindex, i.e., with a single index and with the
use of the \printindex command.
Then we need to invoke makeindex.
In addition, latexmk assumes that makeindex or its replacement makes
an .ind file from an .idx file, and latexmk gives an error if it
doesn't exist, we need to make an .ind file.
Both problems are solved by running makeindex and then splitindex.
Note: errors are returned by makeindex and splitindex for things
like a missing input file. No error is returned for lines in an
input file that are in an incorrect format; they are simply
ignored. So no problem is caused by lines in the .idx file
that are generated by splitindex in a format incompatible with
makeindex.
my $ret1 = system( "makeindex", $$Psource );
my $ret2 = system( "splitindex", $$Psource );
return $ret1 || $ret2;
}
add_cus_dep( 'acn', 'acr', 0, 'makeglossaries' );
add_cus_dep( 'glo', 'gls', 0, 'makeglossaries' );
push @generated_exts, 'glo', 'gls', 'glg';
push @generated_exts, 'acn', 'acr', 'alg';
push @generated_exts, "ist"; # index stylefile created by the glossaries package
#$clean_ext .= " acr acn alg glo gls glg";# TBD: clarify: better in @generated_exts?
sub makeglossaries {
my ($base_name, $path) = fileparse( $_[0] );
my @args = ( "-q", "-d", $path, $base_name );
if ($silent) { unshift @args, "-q"; }
return system "makeglossaries", "-d", $path, $base_name;
}
sub run_makeglossaries {
my ($base_name, $path) = fileparse( $_[0] ); #handle -outdir param by splitting path and file, ...
pushd $path; # ... cd-ing into folder first, then running makeglossaries ...
if ( $silent ) {
# system "makeglossaries -q '$base_name'"; #unix
system "makeglossaries", "-q", "$base_name"; #windows
}
else {
# system "makeglossaries '$base_name'"; #unix
system "makeglossaries", "$base_name"; #windows
};
popd; # ... and cd-ing back again
}
!!! ONLY WORKS WITH VERSION 4.54 or higher of latexmk
#TBD: take into account: modified:
'$_[0]'->
# makeindex
@ist = glob("*.ist");
if (scalar(@ist) > 0) {
$makeindex = "makeindex -s $ist[0] %O -o %D %S";
}
glossaries new with bib2gls
push @generated_exts, 'glstex', 'glg';
add_cus_dep('aux', 'glstex', 0, 'run_bib2gls');
sub run_bib2gls {
if ( $silent ) {
my $ret = system "bib2gls --silent --group $_[0]";
} else {
my $ret = system "bib2gls --group $_[0]";
};
my ($base, $path) = fileparse( $_[0] );
if ($path && -e "$base.glstex") {
rename "$base.glstex", "$path$base.glstex";
}
# Analyze log file.
local *LOG;
$LOG = "$_[0].glg";
if (!$ret && -e $LOG) {
open LOG, "<$LOG";
while (<LOG>) {
if (/^Reading (.*.bib)\s$/) {
rdb_ensure_file( $rule, $1 );
}
}
close LOG;
}
return $ret;
}
$pythontex = 'pythontexW %R';#'pythontexW %O %R';
push @generated_exts, "pytxcode", "plg";
push @generated_exts, "depytx", "dplg";
$clean_ext .= " pythontex-files-%R/* pythontex-files-%R";
#$extra_rule_spec{'pythontex'} = [ 'internal', '', 'mypythontex', "%Y%R.pytxcode", "%Ypythontex-files-%R/%R.pytxmcr", "%R", 1 ];
$extra_rule_spec{'pythontex'} = [ 'internal', '', 'mypythontex', "%R.pytxcode", "pythontex-files-%R/%R.pytxmcr", "%R", 1 ];
sub mypythontex {
my $result_dir = $aux_dir1."pythontex-files-$$Pbase";
my $ret = Run_subst( $pythontex, 2 );
rdb_add_generated( glob "$result_dir/*" );
#my $fh = new FileHandle $$Pdest, "r";
open( my $fh, "<", $$Pdest );
if ($fh) {
print "path: $ENV{PATH}";
while (<$fh>) {
if ( /^%PythonTeX dependency:\s+'([^']+)';/ ) {
print "Found pythontex dependency '$1'\n";
rdb_ensure_file( $rule, $aux_dir1.$1 );
}
}
undef $fh;
}
else {
warn "mypythontex: I could not read '$$Pdest'\n",
" to check dependencies\n";
}
return $ret;
}
for htlatex
push @generated_exts, "4tc", "4ct", "tmp", "xref", "css", "idv", "lg";
TBD: for -C remove also html and xhtml
TBD: check that this plugin also removes all these extensions.. think of lg.
biblatex
push @generated_exts, "run.xml";# does run.xml work?
$clean_ext .= " %R-blx.bib";
What else is your idea?
– user2609605 May 13 '23 at 21:47\customlstinputlistingetc. and use it instead of the original\lstinputlisting. – user202729 May 14 '23 at 23:52