pdfjam

David Firth https://warwick.ac.uk/dfirth

A snapshot of this page is included as README.md in the distributed package. The most up-to-date version is maintained online at https://github.com/DavidFirth/pdfjam.

Overview What is pdfjam?

The pdfjam package makes available the pdfjam shell script that provides a simple interface to much of the functionality of the excellent pdfpages package (by Andreas Matthias) for . The pdfjam script takes one or more PDF files (and/or JPG/PNG graphics files) as input, and produces one or more PDF files as output. It is useful for joining files together, selecting pages, reducing several source pages onto one output page, etc., etc.

A potential drawback of pdfjam and other scripts based upon it is that any hyperlinks in the source PDF are lost. On the positive side, there is no appreciable degradation of image quality in processing PDF files with pdfjam, unlike some other indirect methods such as pdf2ps|psnup|ps2pdf (in the author's experience).

pdfjam is designed for Unix-like systems, including Linux and Mac OS X. It seems that it will work also on Windows computers with a suitable installation of Cygwin (with Live installed), but this has not been thoroughly tested.

An alternative set of PDF manipulation tools, which are java-based, is provided by the Multivalent project. Yet another alternative set of tools is PDFsam. Those alternatives do much the same things as pdfjam, and maybe quite a bit more too.

The pdfjam software is made available free, under GPL version 2 (see the file named COPYING that is included with the package). It comes with ABSOLUTELY NO WARRANTY of fitness for any purpose whatever.

Wrapper scripts no longer included here

Previous versions of pdfjam (before 3.02) included some other scripts too, in addition to the pdfjam script iteslf. Those other scripts are simple wrappers for calls to pdfjam, designed to perform some common tasks such as joining or n-upping PDF files or to illustrate other features; they are not very elaborate, and nor are they extensively tested. They are probably best viewed as simple templates that can be used for constructing more elaborate wrapper scripts as required.

Those 'wrapper' scripts are no longer maintained. I continue to make them available in a separate repository https://github.com/DavidFirth/pdfjam-extras, in case anyone wants to see them, to use them, or to improve and maintain them independently as a separate project.

The specific wrapper scripts that were removed from the pdfjam package at version 3.02 are:

• pdfnup, pdfpun
• pdfjoin
• pdf90, pdf180, pdf270
• pdfflip
• pdfbook
• pdfjam-pocketmod
• pdfjam-slides3up, pdfjam-slides6up

Pre-requisites
• A Unix-like operating system (Linux, Mac OS X, FreeBSD, etc.; possibly Cygwin)
• A working, up-to-date installation of (including at least one of pdflatex, lualatex or xelatex)
• A working installation of the package pdfpages (version 0.4f or later)
• Some of the options offered by pdfjam also require the packages geometry and hyperref.

For some years now, pdfjam has been included in the Live distribution, which includes all the necessary programs and packages to make pdfjam run smoothly. If you have the necessary bandwidth and disk space for it, I do recommend installing Live.

Documentation

The primary documentation for pdfjam is obtained (after installation of pdfjam) via the command

pdfjam --help

This gives information on the arguments to pdfjam, and the default settings that apply at your installation. In addition to the arguments that are explicitly documented there, pdfjam provides access to all of the options of the pdfpages package: that's a large number of options, and it's a set of options that might change, so users are referred to the current pdfpages manual (PDF) to see what's available.

In case you want to look at the pdfjam --help text template online, for example before you decide whether to install pdfjam, the source is available at https://github.com/DavidFirth/pdfjam/blob/master/pdfjam-help.txt.

There is also a (very basic) man page, accessed in the usual way (after installation) by

man pdfjam

Installation

There are two main ways:

1. Install the current Live distribution

(how you do this will depend on details of your operating system). Live will already contain a recent release of pdfjam. Many thanks to Karl Berry for setting up and maintaining pdfjam as a CTAN package that is part of Live.

1. Install it yourself (e.g., if you don't want Live, or if you want a later

release of pdfjam than the one that's currently in Live). Download the latest packaged release of pdfjam from https://github.com/DavidFirth/pdfjam/releases. If for some reason you don't want the latest released version, or even a recently released version, you can still get older versions too: see https://davidfirth.github.io/pdfjam.

The first way, via Live, is recommended as the easiest way for most users.

If you go the second way, then you will have a bit more to do:

• The pdfjam shell script in the bin sub-directory of the released package should be placed on the PATH of anyone who needs to use it.
• The man file in the man1 sub-directory should be installed on the MANPATH of all who might need to read it.

Configuration

On many unix-like systems pdfjam should run without any further configuration, provided that the pre-requisite installation (such as Live) is present. If you want to check (e.g., prior to installation) that pdfjam will work on your system, then

• unzip the tests.zip archive (inside the package)
• cd to your newly made tests sub-directory and follow the instructions that appear there in README_tests.md.

If configuration is needed, this can be done through a site-wide or user-specific configuration file. This might be necessary if, for example, your site has a non-standard installation, or a non-standard location for temporary files, or a specific paper size for output PDFs — or some other reason.

The file pdfjam.conf is a sample configuration file which can be edited as needed. After editing, either install the file for site-wide use (at /etc/pdfjam.conf, /usr/share/etc/pdfjam.conf, /usr/local/share/pdfjam.conf, or /usr/local/etc/pdfjam.conf) or as a user-defaults file at ~/.pdfjam.conf. User settings made at ~/.pdfjam.conf override corresponding settings made in a site-wide configuration file.

For example, if you prefer to use /usr/bin/xelatex as your default engine (in place of the standard pdflatex), and you want your output page size to be "US letter" size paper by default, you would simply include the lines

latex=/usr/bin/xelatex
paper=letterpaper

in a plain text file named '.pdfjam.conf' in your home directory. The path usr/bin/xelatex specifies exactly which program will be used: it could be a path to any one of the pdflatex, lualatex or xelatex executables. (You can get the full path to an executable by, for example, the command which xelatex.) The code word letterpaper is how refers to that particular page size. For other available paper sizes, and all the many other options that could be set as defaults if you want, please see the output of

pdfjam --help

On some systems it might even be necessary to change the list of places (i.e., /etc/pdfjam.conf and others as listed above) that is searched for site-wide configuration files. This can only be done by editing the pdfjam script itself. To see which directories on your system are searched for a file named pdfjam.conf, look at the output of

pdfjam --configpath

Using pdfjam

For a full overview of what pdfjam can do, the importance of the pdfpages manual (PDF) cannot be stressed enough! The following examples merely serve as a brief introduction.

Example 1: Batch 2-upping of documents

Consider converting each of two documents to a side-by-side "2-up" format. Since we want the two documents to be processed separately, we'll use the --batch option:

pdfjam --batch --nup 2x1 --suffix 2up --landscape file1.pdf file2.pdf

This will produce new files file1-2up.pdf and file2-2up.pdf in the current working directory.

Example 2: Merging pages from 2 documents

Suppose we want a single new document which puts together selected pages from two different files:

pdfjam file1.pdf '{},2-' file2.pdf '10,3-6' --outfile ../myNewFile.pdf

The new file myNewFile.pdf, in the parent directory of the current one, contains an empty page, followed by all pages of file1.pdf except the first, followed by pages 10, 3, 4, 5 and 6 from file2.pdf.

The resulting PDF page size will be whatever is the default paper size for you at your site. If instead you want to preserve the page size of (the first included page from) file1.pdf, use the option --fitpaper true.

All pages in an output file from pdfjam will have the same size and orientation. For joining together PDF files while preserving different page sizes and orientations, pdfjam is not the tool to use.

Example 3: A 4-up document with frames

To make a portrait-oriented 4-up file from the pages of three input files, with a thin-line frame around the input pages:

pdfjam file1.pdf file2.pdf file3.pdf --no-landscape --frame true --nup 2x2 --suffix 4up --outfile ~/Documents

Here a directory was specified at --outfile: the resultant file in this case will be ~/Documents/file3-4up.pdf. (Note that if there's a writeable file with that name already, it will be overwritten: no check is made, and no warning given.)

Example 4: Convert a 'US letter' document to A4

Suppose we have a document made up of 'US letter' size pages, and we want to convert it to A4:

pdfjam 'my US letter file.pdf' --a4paper --outfile 'my A4 file.pdf'

Example 5: Handouts from presentation slides

A useful application of pdfjam is for producing a handout from a file of presentation slides. For slides made with the standard 4:3 aspect ratio a nice 6-up handout on A4 paper can be made by

pdfjam --nup 2x3 --frame true --noautoscale false --delta "0.2cm 0.3cm" --scale 0.95 myslides.pdf --outfile myhandout.pdf

The --delta option here comes from the pdfpages package; the --scale option is passed to 's \includegraphics command.

Slides made by 's beamer package, using the handout class option, work especially nicely with this! The example wrapper scripts pdfjam-slides3up and pdfjam-slides6up, in the pdfjam-extras repository, are for 3-up and 6-up handouts, respectively.

Example 6: Trimming pages; and piped output

Suppose we want to trim the pages of our input file prior to n-upping. This can be done by using a pipe:

pdfjam myfile.pdf --trim '1cm 2cm 1cm 2cm' --clip true --outfile /dev/stdout | pdfjam --nup 2x1 --frame true --outfile myoutput.pdf

The --trim option specifies an amount to trim from the left, bottom, right and top sides respectively; to work as intended here it needs also --clip true. These (i.e., trim and clip) are in fact options to 's \includegraphics command (in the standard graphics package).

Thanks go to Christophe Lange and Christian Lohmaier for suggesting an example on this.

Example 7: Output pages suitable for binding

To offset the content of double-sided printed pages so that they are suitable for binding with a Heftstreifen, use the --twoside option:

pdfjam --twoside myfile.pdf --offset '1cm 0cm' --suffix 'offset'

Example 8: Input file with nonstandard name

To use PDF input files whose names do not end in '.pdf', you will need to use the --checkfiles option. This depends on the availability of the file utility, with support for the options -Lb; this can be checked by trying

file -Lb 'my PDF file'

where 'my PDF file' is the name of a PDF file on your system. The result should be something like 'PDF document, version 1.4' (possibly with a different version number).

With 'file -Lb' available, we can use PDF files whose names lack the usual '.pdf' extension. For example,

pdfjam --nup 2x1 --checkfiles 'my PDF file'

will result in a file named 'my PDF file-2x1.pdf' in the current working directory.

Example 9: Rotate every 2nd page

If you want to print a landscape-oriented PDF document on both sides of the paper, using a duplex printer that does not have 'tumble' capability, make a new version with every second page rotated for printing:

pdfjam --landscape --doublepagestwistodd true my-landscape-document.pdf

FAQ

1. The script runs but the output doesn't look the way it should. Why?

Most likely either your pdf or your pdfpages installation is an old version. You could check also that pdftex.def, typically to be found in .../texmf/tex/latex/graphics/, is up to date. If the problem persists even with up-to-date versions of pdf, pdftex.def and pdfpages, then please do report it.

2. What can I do to solve a 'Too many open files' error?

This error has been reported to occur sometimes, when dealing with large numbers of documents/pages. A suggested solution, if this happens, is to include additionally (in the call to pdfjam):

    -- preamble '\let\mypdfximage\pdfximage \def\pdfximage{\immediate\mypdfximage}'

Reporting bugs

Please report any bugs found in pdfjam, either

Some known problems:

• Sometimes font information (such as ligatures) is lost from the output of pdfjam. It seems that a fairly simple fix when this happens is to add the option --preamble '\pdfinclusioncopyfonts=1' in your call to pdfjam.
• In Cygwin, using pdfjam in a pipeline does not seem to work. The problem seems to be with Cygwin's handling of file descriptors within pipelines.
• The --preamble option can sometimes clash with other elements of the preamble. Some specific things to watch out for:
• If the preamble needs to set further options to the geometry package, when the geometry package has already been loaded in order to set a special page size, be sure to use the \geometry{} command for that.
• If the preamble needs to set options to the hyperref package, when
the hyperref package has already been loaded in order to specify PDF document information, be sure to use the \hypersetup{} command for that.
• If options to the color package are to be specified in the preamble,
do not at the same time use the --pagecolor option to pdfjam.
Version history

Overview of the history

This all grew originally from a script named pdfnup. That was later joined, in a published package called 'PDFjam', by two further scripts pdfjoin and pdf90.

At version 2.00, everything was unified through a single script pdfjam, with many more options. Along with pdfjam various 'wrapper' scripts — i.e., other scripts that use pdfjam in different ways — were provided, mainly as examples.

From version 3.02, the extra 'wrapper' scripts are removed from the package, mainly because they are hard to maintain: different users want different things, and pdfjam itself provides all the options in any case. So I have broken out the wrapper scripts into a separate repository, unsupported — so that people can still see and use/adapt them if they want. And maybe even someone else will want to take on the task of improving and maintaining some of them, who knows? The wrapper scripts (no longer maintained) can now be found at https://github.com/DavidFirth/pdfjam-extras.

Version release notes

3.03

• Built package now (again) has tests in a zip archive. (needed for CTAN)

3.02

• Re-styled the package name to pdfjam. (Previously the package — which then contained several scripts — was named PDFjam).
• Moved everything to a new home on the web atgithub.com/DavidFirth .
• Simplified the package by removal of all the 'wrapper' scripts (pdfnup, pdfjoin, etc.).
• Settings specified via --preamble are now protected from the normal tidying of the input file to remove redundant packages.
• The possibility to use lualatex or xelatex in place of pdflatex is now described explicitly in the README. The specification of which engine to use can be made either in a configuration file, or on the command line via option (for example) --latex /usr/bin/xelatex. Thanks to Mircea for suggesting this.
• Added new --runs option, so as to allow for example --runs 2 in a situation where two runs of pdflatex (or lualatex or xelatex) are needed (typically where the result document is being indexed in some way, so more than one run is needed). Thanks to Ferdinand for this suggestion (and patch).
• Default paper size is now guessed from the locale where possible, with fallback default size being ISO A4 ('a4paper' in ). This can still be over-ridden in a configuration file, or on the command line. Thanks to Jonathan for this suggestion.
• Innocuous edits made to make pdfjam work better in Cygwin. Thanks to Lucas for sending a helpful patch for this.
• Included a note in the FAQ about the (sometimes reported) 'Too many open files' error. Thanks to George for information about this.
• Tidying of the pdfjam script, and better organisation of the tests folder. Thanks to Lucas for help with this.

———-

Older releases are still available at https://davidfirth.github.io/pdfjam.

2.09

• The default behaviour of pdfbook is reverted to its pre-2.06 state, because --booklet true seems to be problematic for some users.

2.08

• Fixed a bug in one of the tests.

2.07

• Two other common graphics formats (JPG and PNG) are now explicitly allowed as input files (i.e., not only PDF files are allowed as inputs).

2.06

• Changed the pdfbook script to include --booklet true as the default behaviour (thanks to Julien Bossert for this good suggestion).

2.05

• Changes to the pdfbook script [the --right-edge-binding option is now redundant, and there's a new --short-edge option for binding along the short edge of pages instead of the long edge (thanks to Marco Pessotto for this).
• The --preamble option to pdfjam is enhanced, to allow multiple instances which get concatenated.
• Also various minor corrections to man pages.

2.04

• Various minor improvements suggested by Debian maintainers (thanks to Eduard Bloch for these). The main things are:
• addition of the --version option;
• liberalisation of pdfjam to allow files in JPEG format to be specified as input, as well as PDF;
• tidying of the man files;
• and more use of exec, to avoid forking.

2.03

• Fixed a bug which caused problems when your /bin/sh is the zsh shell.
• Fixed a bug which prevented the correct representation of many UTF-8 characters in pdfinfo data.

2.02

• More progress on portability.
• Introduced the beginnings of a suite of tests.
• Iimprovements in the --keepinfo functionality, and in the treatment of file permissions (thanks to Marco Pessotto for these).

2.01

• Fixed a silly bug (thoughtless use of "test -a" in a couple of places) which seriously affected portability.

2.00 major re-design

This is not completely backward-compatible with previous versions of the pdfnup, pdfjoin and pdf90 scripts. The differences in interface are few, though, and the main ones are listed below.

• The new script pdfjam now does all the work; all the other scripts included with the package are just simple wrappers for pdfjam.
• New pdfjam makes available essentially all of the facilities of the pdfpages package (without having to know what they are).
• Various security and portability issues have been resolved. (None of the scripts now calls for /bin/bash; and the handling of temporary files is now much safer.)
• pdfjam can take PDF input from /dev/stdin, and send output to /dev/stdout. (This allows pdfjam to be used in a pipeline.)
• If the --outfile option specifies a relative path, that path is now relative to the current working directory (as is normally expected of unix utilities). This is different behaviour from previous versions.
• With multiple input files, pdfjam offers two distinct methods of processing. This is different behaviour from previous versions.
1. The default is to take pages as specified from the input files, and
combine them into a single document.
1. If the --batch option is used, pdfjam operates separately on the
input files, producing one output file per input file.
• Page selection is available separately for each input file. A difference from previous versions is that the --pages keyword is no longer used.
• pdfjam uses the \includepdfmerge command from pdfpages, as a result of which all pages in an output file have the same size and orientation. This is different behaviour from previous versions.
• Output page orientation is now controlled by using --landscape (negated, if necessary, by --no-landscape). The previous --orient option is no longer used, and in particular "--orient auto" is no longer available. This is different behaviour from previous versions.
• Other new features include:
• --keepinfo option, to allow preservation of PDF document information, if the pdfinfo utility and the hyperref package are available (thanks to Robert Wenner for suggesting this);
• --pdftitle, --pdfauthor, --pdfsubject and --pdfkeywords
options, to specify new PDF document information (these require the hyperref package);
• many more "named" output page sizes are available if the
geometry package is installed (the full list of allowed paper specifications is: a0paper, a1paper, a2paper, a3paper, a4paper, a5paper, a6paper b0paper, b1paper, b2paper, b3paper, b4paper, b5paper, b6paper letterpaper, executivepaper, legalpaper; thanks to Corné Verbruggen and Mel Irizarry for suggesting this) and non-standard page sizes can also be defined;
• --checkfiles option to request that input files be checked using the
file utility, rather than requiring the file name to end in ".pdf" or ".PDF";
• --twoside option, to allow the twoside class option to be
specified (thanks to Johannes Reinhard for suggesting this);
• --pagecolor option, to allow the background colour of output pages
to be changed (thanks to James Fisher for suggesting this);
• --vanilla option to run pdfjam without reading site-wide or user
configuration files.
• In addition, various reported bugs have been fixed — many thanks to all those kind people who reported them.

1.21

• Bug fixes, including security issues (many thanks to Eduard Bloch, Robert Buchholz and Martin Vaeth for helpful reports on vulnerabilities and for kindly contributing patches).
• The scripts now call for /bin/bash as interpreter.
• Availability of mktemp is now also assumed.
• The Mac OS X droplets now look for pdflatex at (by default) /usr/texbin/pdflatex

1.20 2005-01-25

• Added extra possible locations for the site-wide configuration file.

1.11

• Added the --scale option to pdfnup, which allows page margins either to be enlarged (e.g,. --scale 0.9) or reduced (e.g,. --scale 1.1) by scaling the page contents. By popular request!

1.10

• Output files now appear by default in same directory as input, rather than in the current working directory; fixed a bug that caused the scripts not to work on some versions of Solaris (thanks to Daniel Gebhart); major improvements to the Mac OS X sample droplets.

1.03

• Minor changes towards POSIX compliance.

1.02

• Added a COPYING file to the package.

1.00

• Package re-named PDFjam.

0.99a

• A minor change to the output of pdfnup --help and pdfjoin --help

0.99

• Various improvements to pdfnup, including the handling of multiple PDF

input files. Added pdfjoin and pdf90.

0.97

• Corrections to the output of pdfnup --help.

0.96

• Minor changes to comments in the pdfnup script.

0.95

• Added the possibility of site-specific and user-specific configuration files (thanks to Jason Lewis for suggesting this).

0.9

• Added --openright (thanks to Jason Lewis for suggesting this).

0.8

• Added pdfnup --help facility (thanks to Wilfrid Kendall for this suggestion).

0.7

• Paths involving spaces now permitted.
• Page trimming added (thanks to Alex Montgomery for suggesting that).
• Default output filename now has a dash inserted before the "nup" label (as in  wasteful-2x2.pdf ).
• Sample Mac OS X droplets provided.

0.6

• Use of paths involving spaces now reports an error.

0.5

• Fixed a bug which caused incompatibility with some types of unix.

0.4

• Better error trapping, improved portability.

0.3

• First public release of pdfnup.