Tools to parse and evaluate the Puppet DSL.

Version on this page:
LTS Haskell 16.31:
Stackage Nightly 2020-06-16:
Latest on Hackage:1.5.1

See all snapshots language-puppet appears in

BSD-3-Clause licensed by Simon Marechal
Maintained by [email protected]
This version can be pinned in stack with:language-puppet-,10336

Module documentation for

= Language-puppet

image:["Build Status", link=""]

A library to work with Puppet manifests, test them and eventually replace everything ruby.

.Basic usage:
puppetresources -p /where/your/puppet/files/are -o

.Quick install instruction with stack
stack install language-puppet

.Build instructions:
git clone
cd language-puppet
# Using nix
cabal2nix . > default.nix
nix-build shell.nix
# Using stack
ln -s stack-10.yaml stack.yaml
stack build

== Puppetresources

The `puppetresources` command is a command line utility that let you interactively compute catalogs on your local computer.
It is much faster than its ruby counterpart, and has been designed for giving assistance to the Puppet catalog writer.

There are 4 different modes:

* `--node` will display all resources on screen in a nice user-friendly colored fashion.
* `--all` displays statitics and optionally shows dead code.
* `--parse` only goes as far as parsing. No interpretation.
* `--showcontent` to display file content.

Catalog is not computed exactly the same way Puppet does. Some good practices are enforced. A strict and more permissive mode are provided.

=== Command line arguments

`-p` or `--puppetdir`::

This argument is mandatory except in `parse` mode. It must point to the base of the puppet directory (the directory that contains the `modules` and `manifests` directories).

`-o` or `--node`::

Enable the `node mode`. This let you specify the name of the node you wish to compute the catalog for.

`-a` or `-all`::

Enable the `stats mode`. If you specify `allnodes` it will compute the catalogs for all nodes that are specified in `site.pp` (this will not work for regexp-specified or the default nodes). You can also specify a list of nodes separated by a comma.
Combined with `--deadcode`, it will display the list of puppet files that have not been used.
This is useful as automated tests, to check a change didn't break something. You might want to run this option with `+RTS -N`.

`-t` or `--type`::

Filters the resources of the resulting catalog by type. Using PCRE regex is supported.

`-n` or `--name`::

Filters the resources of the resulting catalog by name. Using PCRE regex is supported.

`-c` or `--showcontent`::

If `-n` is the exact name of a file type resource defined in the catalog, this will display the file content nicely. Useful for debugging templates.
Example: `puppetresources -p . -o mynodename -n '/etc/motd' --showcontent`

`--loglevel` or `-v`::

Possible values are : DEBUG, INFO, NOTICE, WARNING, ERROR


Expects the url of a live PuppetDB.


Expects a path to a *fake* PuppetDB, represented as a YAML file on disk. This option is pretty slow but can be invaluable to test exported resources tricks.


Expects the path to the `hiera.yaml` file.


Expects a list of comma-separated modules. The interpreter will not try to parse and evaluate the defined types and classes from this module. This is useful for using modules that use bad
practices forbidden by `puppetresources`.


When this flag is set, exported resources, catalogs and facts are saved in the PuppetDB. This is useful in conjunction with `--pdbfile`.


When this flag is set, exported resources are saved in the PuppetDB. This is useful in conjunction with `--pdbfile`.

`-j` or `--JSON`::

Displays the catalog as a Puppet-compatible JSON file, that can then be used with `puppet apply`.


Enable strict check.
Strict is less permissive than vanilla Puppet.
It is meant to prevent some pitfalls by enforcing good practices.
For instance it refuses to
- silently ignore/convert `undef` variables
- lookup an hash with an unknown key and return `undef`.


Disable the extra tests from `Puppet.OptionalTests`.


Enable `parse mode`. Specify the puppet file to be parsed. Variables are not resolved. No interpretation.


Output version information and exist.

=== Settings defaults using a yaml file

Defaults for some of these options can be set using a `/yourworkingdirectory/tests/defaults.yaml` file. For instance `OptionalTests` is checking that all users and groups are known. Because some of these users and groups might be defined outside puppet, a list of known ones is used internally. This can be overridden in that file using the key `knownusers` and `knowngroups`.

Please look at[the template file] for a list of possible defaults.

== pdbQuery

The `pdbquery` command will work with different implementations of PuppetDB (the official one with its HTTP API, the file-based backend and dummy ones). It can be used to:

* export data from production PuppetDB to a file (in order to debug some issue with `puppetresources`).
* query a Puppetdb

Here is a list of command line arguments :

`-l` or `--location`::

The URL of the PuppetDB when working with a remote PuppetDB, a file path when working with the file-based test implementation.

`-t` or `--pdbtype`::

The type of PuppetDB to work with:

* dummy: a dummy PuppetDB.
* remote: a "real" PuppetDB, accessed by its HTTP API.
* test: a file-based backend emulating a PuppetDB.


Output facts for a specific node (json)

Output all nodes (json)

Output all resources for a specific node (json)

Dump all facts to `/tmp/allfacts.yaml`.

Create a test DB from the current DB

Adds facts to the test DB for the given node name, if they are not already defined.

Output version information and exit.

== Unsupported Puppet idioms or features

* `OS X` is supported when using the latest `filecache-0.3`.

puppet functions::
* the `require` function is not supported (see[issue #17])
* the deprecated `import` function is not supported
* the deprecated node inheritance feature is not supported

custom ruby functions::