Puppet Forge

Access and manipulate the Puppet Forge API from Ruby.

Installation

Add this line to your application's Gemfile:

gem 'puppet_forge'

And then execute:

$ bundle

Or install it yourself as:

$ gem install puppet_forge

Requirements

• Ruby >= 2.6.0

Dependencies

• Faraday 2.x
• Typhoeus ~> 1.0.1 (optional)

Typhoeus will be used as the HTTP adapter if it is available, otherwise Net::HTTP will be used. We recommend using Typhoeus for production-level applications.

Usage

First, make sure you have imported the Puppet Forge gem into your application:

require 'puppet_forge'

Next, supply a user-agent string to identify requests sent by your application to the Puppet Forge API:

PuppetForge.user_agent = "MyApp/1.0.0"

Now you can make use of the resource models defined by the gem:

• PuppetForge::V3::User
• PuppetForge::V3::Module
• PuppetForge::V3::Release

For convenience, these classes are also aliased as:

• PuppetForge::User
• PuppetForge::Module
• PuppetForge::Release

The aliases will always point to the most modern API implementations for each model. You may also use the fully qualified class names (e.g. PuppetForge::V3::User) to ensure your code is forward compatible.

See the Basic Interface section below for how to perform common tasks with these models.

Please note that PuppetForge models are identified by unique slugs rather than numerical identifiers.

The slug format, properties, associations, and methods available on each resource model are documented on the Resource Reference page.

Basic Interface

Each of the models uses ActiveRecord-like REST functionality to map over the Forge API endpoints. Most simple ActiveRecord-style interactions function as intended.

Currently, only read-only actions are supported.

The methods find, where, and all immediately make one API request.

# Find a Resource by Slug
PuppetForge::User.find('puppetlabs') # => #<PuppetForge::V3::User(/v3/users/puppetlabs)>
PuppetForge::Module.find('puppetlabs-stdlib') # => #<PuppetForge::V3::Module(/v3/modules/puppetlabs-stdlib)>

# Find All Resources
PuppetForge::Module.all # See "Paginated Collections" below for important info about enumerating resource sets.

# Find Resources with Conditions
PuppetForge::Module.where(query: 'apache') # See "Paginated Collections" below for important info about enumerating resource sets.
PuppetForge::Module.where(query: 'apache').first # => #<Forge::V3::Module(/v3/modules/puppetlabs-apache)>

For compatibility with older versions of the puppet_forge gem, the following two methods are functionally equivalent.

PuppetForge::Module.where(query: 'apache')
PuppetForge::Module.where(query: 'apache').all # This method is deprecated and not recommended

Errors

All API Requests (whether via find, where, or all) will raise a Faraday::ResourceNotFound error if the request fails.

Downloading and installing a module release

A release tarball can be downloaded and installed by following the steps below.

release_slug = "puppetlabs-apache-1.6.0"
release_tarball = release_slug + ".tar.gz"
dest_dir = "/path/to/install/directory"
tmp_dir = "/path/to/tmpdir"

# Fetch Release information from API
# @raise Faraday::ResourceNotFound error if the given release does not exist
release = PuppetForge::Release.find release_slug

# Download the Release tarball
# @raise PuppetForge::ReleaseNotFound error if the given release does not exist
release.download(Pathname(release_tarball))

# Verify the MD5
# @raise PuppetForge::V3::Release::ChecksumMismatch error if the file's md5 does not match the API information
release.verify(Pathname(release_tarball))

# Unpack the files to a given directory
# @raise RuntimeError if it fails to extract the contents of the release tarball
PuppetForge::Unpacker.unpack(release_tarball, dest_dir, tmp_dir)

Uploading a module release

You can upload new module versions to the forge by following the steps below.

Note: This API requires authorization. See Authorization for more information.

release_tarball = 'pkg/puppetlabs-apache-1.6.0.tar.gz'

# Upload a module tarball to the Puppet Forge
# Returns an instance of V3::Release class and the response from the forge upload 
# @raise PuppetForge::ReleaseForbidden if a 403 response is recieved from the server
# @raise PuppetForge::ReleaseBadContent if the module to upload is not valid
# @raise Faraday::ClientError if any errors encountered in the upload
# @raise PuppetForge::FileNotFound if the given tarball cannot be found
release, response = PuppetForge::V3::Release.upload(release_tarball)

Paginated Collections

The Forge API only returns paginated collections as of v3.

PuppetForge::Module.all.total # => 1728
PuppetForge::Module.all.length # => 20

Movement through the collection can be simulated using the limit and offset parameters, but it's generally preferable to leverage the pagination links provided by the API. For convenience, pagination links are exposed by the library.

PuppetForge::Module.all.offset # => 0
PuppetForge::Module.all.next_url # => "/v3/modules?limit=20&offset=20"
PuppetForge::Module.all.next.offset # => 20

An enumerator exists for iterating over the entire (unpaginated) collection. Keep in mind that this will result in multiple calls to the Forge API.

PuppetForge::Module.where(query: 'php').total # => 37
PuppetForge::Module.where(query: 'php').unpaginated # => #<Enumerator>
PuppetForge::Module.where(query: 'php').unpaginated.to_a.length # => 37

Associations & Lazy Attributes

Associated models are accessible in the API by property name.

PuppetForge::Module.find('puppetlabs-apache').owner # => #<Forge::V3::User(/v3/users/puppetlabs)>

Properties of associated models are then loaded lazily.

user = PuppetForge::Module.find('puppetlabs-apache').owner

# This does not trigger a request
user.username # => "puppetlabs"

# This *does* trigger a request
user.created_at # => "2010-05-19 05:46:26 -0700"

Configuration

To overwrite the default of https://forgeapi.puppet.com and set a custom url for the Forge API:

PuppetForge.host = "https://your-own-api.url/"

Authorization

To authorize your requests with an API key from a Forge user account:

PuppetForge::Connection.authorization = "<your-api-key-here>"

You can generate API keys on your user profile page once you've logged into the Forge website.

Caveats

This library currently does no response caching of its own, instead opting to re-issue every request each time. This will be changed in a later release.

Reporting Issues

Please report problems, issues, and feature requests on the public Puppet Labs issue tracker under the FORGE project. You will need to create a free account to add new tickets.

Contributing

1. Fork it ( https://github.com/[my-github-username]/forge-ruby/fork )
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create a new Pull Request

Releasing

1. Run the release_prep GitHub Action with the new version number, for example, 1.2.3. It will create a new pull request
2. Please check the changelog and ensure all issues/prs have correct labels
3. Run the GitHub Actions release workflow, and it will publish a new version on RubyGems

Contributors

• Pieter van de Bruggen, Puppet Labs
• Jesse Scott, Puppet Labs
• Austin Blatt, Puppet Labs
• Adrien Thebo, Puppet Labs
• Anderson Mills, Puppet Labs

Maintenance

Tickets: File at https://tickets.puppet.com/browse/FORGE