# Installing pandoc
The simplest way to get the latest pandoc release is to use the installer.
Download the latest installer
For alternative ways to install pandoc, see below
under the heading for your operating system.
## Windows
There is a package installer at pandoc's [download page].
This will install pandoc, replacing older versions, and
update your path to include the directory where pandoc's
binaries are installed.
If you prefer not to use the msi installer, we also provide
a zip file that contains pandoc's binaries and
documentation. Simply unzip this file and move the binaries
to a directory of your choice.
Alternatively, you can install pandoc using
[Chocolatey](https://chocolatey.org):
choco install pandoc
Chocolatey can also install other software that integrates with Pandoc.
For example, to install `rsvg-convert` (from [librsvg], covering formats
without SVG support), [Python] (to use Pandoc filters), and
[MiKTeX] (to typeset PDFs with [LaTeX]):
choco install rsvg-convert python miktex
By default, Pandoc creates PDFs using LaTeX.
We recommend installing it via [MiKTeX].
## macOS
There is a package installer at pandoc's [download page].
If you later want to uninstall the package, you can do so
by downloading [this script][uninstaller]
and running it with `perl uninstall-pandoc.pl`.
Alternatively, you can install pandoc using
[Homebrew](http://brew.sh):
brew install pandoc
To include pandoc's citation parser:
brew install pandoc-citeproc
Homebrew can also install other software that integrates with Pandoc.
For example, to install [librsvg] (its `rsvg-convert` covers formats
without SVG support), [Python] (to use Pandoc filters), and
[BasicTeX] (to typeset PDFs with [LaTeX]):
brew install librsvg python homebrew/cask/basictex
Note: On unsupported versions of macOS (more than three releases old),
Homebrew installs from source, which takes additional time and disk space
for the `ghc` compiler and dependent Haskell libraries.
We also provide a zip file containing the binaries and man
pages, for those who prefer not to use the installer. Simply
unzip the file and move the binaries and man pages to
whatever directory you like.
By default, Pandoc creates PDFs using LaTeX. Because a full [MacTeX]
installation uses four gigabytes of disk space, we recommend
[BasicTeX] or [TinyTeX](https://yihui.name/tinytex/)
and using the `tlmgr` tool to install additional packages
as needed. If you receive errors warning of fonts not found:
tlmgr install collection-fontsrecommended
## Linux
Check whether the pandoc version in your package manager is
not outdated. Pandoc is in the [Debian], [Ubuntu], [Slackware],
[Arch], [Fedora], [NiXOS], [openSUSE], [gentoo] and [Void] repositories.
To get the latest release, we provide a binary package for amd64
architecture on the **[download page]**.
This provides both `pandoc` and `pandoc-citeproc`.
The executables are statically linked and
have no dynamic dependencies or dependencies on external
data files. Note: because of the static
linking, the pandoc binary from this package cannot use lua
filters that require external lua modules written in C.
Both a tarball and a deb installer are provided. To install the deb:
sudo dpkg -i $DEB
where `$DEB` is the path to the downloaded deb. This will
install the `pandoc` and `pandoc-citeproc` executables and
man pages.
If you use an RPM-based distro, you may be able to install
the deb from our download page using `alien`.
On any distro, you may install from the tarball into `$DEST`
(say, `/usr/local/` or `$HOME/.local`) by doing
tar xvzf $TGZ --strip-components 1 -C $DEST
where `$TGZ` is the path to the downloaded zipped tarball.
For Pandoc versions before 2.0, which don't provide
a tarball, try instead
ar p $DEB data.tar.gz | tar xvz --strip-components 2 -C $DEST
You can also install from source, using the
instructions below under [Compiling from source].
Note that most distros have the Haskell platform in their
package repositories. For example, on Debian/Ubuntu,
you can install it with `apt-get install haskell-platform`.
For PDF output, you'll need LaTeX. We recommend installing
[TeX Live](http://www.tug.org/texlive/) via your package
manager. (On Debian/Ubuntu, `apt-get install texlive`.)
## Chrome OS
On Chrome OS, pandoc can be installed using the
[chromebrew](https://github.com/skycocker/chromebrew) package manager
with the command:
```sh
crew install pandoc
```
This will automatically build and configure pandoc for the specific
device you are using.
## BSD
Pandoc is in the [NetBSD] and [FreeBSD ports] repositories.
## Docker
The official Docker images for pandoc can be found at
and at
[dockerhub](https://hub.docker.com/).
The [pandoc/core](https://hub.docker.com/r/pandoc/core)
image contains `pandoc` and `pandoc-citeproc`.
The [pandoc/latex](https://hub.docker.com/r/pandoc/latex)
image also contains the minimal LaTeX installation needed
to produce PDFs using pandoc.
To run pandoc using Docker, converting `README.md` to `README.pdf`:
docker run --rm --volume "`pwd`:/data" --user `id -u`:`id -g` pandoc/latex README.md -o README.pdf
## GitHub Actions
Pandoc can be run through
[GitHub Actions](https://github.com/features/actions). For some
examples, see .
## Compiling from source
If for some reason a binary package is not available for your
platform, or if you want to hack on pandoc or use a non-released
version, you can install from source.
### Getting the pandoc source code
Source tarballs can be found at
. For example, to
fetch the source for version 1.17.0.3:
wget https://hackage.haskell.org/package/pandoc-1.17.0.3/pandoc-1.17.0.3.tar.gz
tar xvzf pandoc-1.17.0.3.tar.gz
cd pandoc-1.17.0.3
Or you can fetch the development code by cloning the repository:
git clone https://github.com/jgm/pandoc
cd pandoc
Note: there may be times when the development code is broken
or depends on other libraries which must be installed
separately. Unless you really know what you're doing, install
the last released version.
### Quick stack method
The easiest way to build pandoc from source is to use [stack][stack]:
1. Install [stack][stack]. Note that Pandoc requires stack >= 1.7.0.
2. Change to the pandoc source directory and issue the following commands:
stack setup
stack install
`stack setup` will automatically download the ghc compiler
if you don't have it. `stack install` will install the
`pandoc` executable into `~/.local/bin`, which you should
add to your `PATH`. This process will take a while, and
will consume a considerable amount of disk space.
### Quick cabal method
1. Install the [Haskell platform]. This will give you [GHC] and
the [cabal-install] build tool. Note that pandoc requires
GHC >= 7.10 and cabal >= 2.0.
2. Update your package database:
cabal update
3. Check your cabal version with
cabal --version
If you have a version less than 2.0, install the latest with:
cabal install cabal-install
4. Use `cabal` to install pandoc and its dependencies:
cabal install pandoc
This procedure will install the released version of pandoc,
which will be downloaded automatically from HackageDB.
If you want to install a modified or development version
of pandoc instead, switch to the source directory and do
as above, but without the 'pandoc':
cabal install
5. Make sure the `$CABALDIR/bin` directory is in your path. You should
now be able to run `pandoc`:
pandoc --help
[Not sure where `$CABALDIR` is?](http://www.haskell.org/haskellwiki/Cabal-Install#The_cabal-install_configuration_file)
5. If you want to process citations with pandoc, you will also need to
install a separate package, `pandoc-citeproc`. This can be installed
using cabal:
cabal install pandoc-citeproc
By default `pandoc-citeproc` uses the "i;unicode-casemap" method
to sort bibliography entries (RFC 5051). If you would like to
use the locale-sensitive unicode collation algorithm instead,
specify the `unicode_collation` flag:
cabal install pandoc-citeproc -funicode_collation
Note that this requires the `text-icu` library, which in turn
depends on the C library `icu4c`. Installation directions
vary by platform. Here is how it might work on macOS with Homebrew:
brew install icu4c
stack install pandoc-citeproc \
--flag "pandoc-citeproc:unicode_collation" \
--extra-lib-dirs=/usr/local/opt/icu4c/lib \
--extra-include-dirs=/usr/local/opt/icu4c/include
6. The `pandoc.1` man page will be installed automatically. cabal shows
you where it is installed: you may need to set your `MANPATH`
accordingly. If `MANUAL.txt` has been modified, the man page can be
rebuilt: `make man/pandoc.1`.
The `pandoc-citeproc.1` man page will also be installed automatically.
### Custom cabal method
This is a step-by-step procedure that offers maximal control
over the build and installation. Most users should use the
quick install, but this information may be of use to packagers.
For more details, see the [Cabal User's Guide]. These instructions
assume that the pandoc source directory is your working directory.
You will need cabal version 2.0 or higher.
1. Install dependencies: in addition to the [Haskell platform],
you will need a number of additional libraries. You can install
them all with
cabal update
cabal install --only-dependencies
2. Configure:
cabal configure --prefix=DIR --bindir=DIR --libdir=DIR \
--datadir=DIR --libsubdir=DIR --datasubdir=DIR --docdir=DIR \
--htmldir=DIR --program-prefix=PREFIX --program-suffix=SUFFIX \
--mandir=DIR --flags=FLAGSPEC --enable-tests
All of the options have sensible defaults that can be overridden
as needed.
`FLAGSPEC` is a list of Cabal configuration flags, optionally
preceded by a `-` (to force the flag to `false`), and separated
by spaces. Pandoc's flags include:
- `embed_data_files`: embed all data files into the binary (default no).
This is helpful if you want to create a relocatable binary.
- `https`: enable support for downloading resources over https
(using the `http-client` and `http-client-tls` libraries).
3. Build:
cabal build
cabal test
4. Build API documentation:
cabal haddock --html-location=URL --hyperlink-source
5. Copy the files:
cabal copy --destdir=PATH
The default destdir is `/`.
6. Register pandoc as a GHC package:
cabal register
Package managers may want to use the `--gen-script` option to
generate a script that can be run to register the package at
install time.
### Creating a relocatable binary
It is possible to compile pandoc such that the data files
pandoc uses are embedded in the binary. The resulting binary
can be run from any directory and is completely self-contained.
With cabal, add `-fembed_data_files` to the `cabal configure`
or `cabal install` commands.
With stack, use `--flag pandoc:embed_data_files`.
### Running tests
Pandoc comes with an automated test suite.
To run with cabal, `cabal test`; to run with stack, `stack
test`.
To run particular tests (pattern-matching on their names), use
the `-p` option:
cabal install pandoc --enable-tests
cabal test --test-options='-p markdown'
Or with stack:
stack test --test-arguments='-p markdown'
It is often helpful to add `-j4` (run tests in parallel)
and `--hide-successes` (don't clutter output with successes)
to the test arguments as well.
If you add a new feature to pandoc, please add tests as well, following
the pattern of the existing tests. The test suite code is in
`test/test-pandoc.hs`. If you are adding a new reader or writer, it is
probably easiest to add some data files to the `test` directory, and
modify `test/Tests/Old.hs`. Otherwise, it is better to modify the module
under the `test/Tests` hierarchy corresponding to the pandoc module you
are changing.
### Running benchmarks
To build and run the benchmarks:
cabal configure --enable-benchmarks && cabal build
cabal bench
or with stack:
stack bench
To use a smaller sample size so the benchmarks run faster:
cabal bench --benchmark-options='-s 20'
To run just the markdown benchmarks:
cabal bench --benchmark-options='markdown'
[Arch]: https://www.archlinux.org/packages/community/x86_64/pandoc/
[Cabal User's Guide]: http://www.haskell.org/cabal/release/latest/doc/users-guide/builders.html#setup-configure-paths
[Debian]: https://packages.debian.org/pandoc
[Fedora]: https://apps.fedoraproject.org/packages/pandoc
[FreeBSD ports]: http://www.freshports.org/textproc/pandoc/
[GHC]: http://www.haskell.org/ghc/
[GPL]: http://www.gnu.org/copyleft/gpl.html
[Haskell platform]: http://hackage.haskell.org/platform/
[MacPorts]: http://trac.macports.org/browser/trunk/dports/textproc/pandoc/Portfile
[MacTeX]: https://tug.org/mactex/
[BasicTeX]: http://www.tug.org/mactex/morepackages.html
[LaTeX]: https://www.latex-project.org
[MiKTeX]: http://miktex.org/
[librsvg]: https://wiki.gnome.org/Projects/LibRsvg
[Python]: https://www.python.org
[NetBSD]: http://pkgsrc.se/wip/pandoc
[NixOS]: https://nixos.org/nixos/packages.html
[Slackware]: https://www.slackbuilds.org/result/?search=pandoc&sv=
[Ubuntu]: https://packages.ubuntu.com/pandoc
[download page]: https://github.com/jgm/pandoc/releases/latest
[gentoo]: http://packages.gentoo.org/package/app-text/pandoc
[haskell repository]: https://wiki.archlinux.org/index.php/Haskell_Package_Guidelines#.5Bhaskell.5D
[openSUSE]: https://software.opensuse.org/package/pandoc
[source tarball]: http://hackage.haskell.org/package/pandoc
[stack]: https://docs.haskellstack.org/en/stable/install_and_upgrade.html
[cabal-install]: http://hackage.haskell.org/package/cabal-install
[Void]: https://voidlinux.org/
[uninstaller]: https://raw.githubusercontent.com/jgm/pandoc/master/macos/uninstall-pandoc.pl