Forge Home

repoutil

Interact with package repositories in OS-indepenent way

12,768 downloads

7,336 latest version

3.8 quality score

We run a couple of automated
scans to help you access a
module's quality. Each module is
given a score based on how well
the author has formatted their
code and documentation and
modules are also checked for
malware using VirusTotal.

Please note, the information below
is for guidance only and neither of
these methods should be considered
an endorsement by Puppet.

Version information

  • 0.1.5 (latest)
  • 0.1.4
  • 0.1.3
  • 0.1.2
  • 0.1.1
  • 0.1.0
released Jun 14th 2016

Start using this module

  • r10k or Code Manager
  • Bolt
  • Manual installation
  • Direct download

Add this module to your Puppetfile:

mod 'ptomulik-repoutil', '0.1.5'
Learn more about managing modules with a Puppetfile

Add this module to your Bolt project:

bolt module add ptomulik-repoutil
Learn more about using this module with an existing project

Manually install this module globally with Puppet module tool:

puppet module install ptomulik-repoutil --version 0.1.5

Direct download is not typically how you would use a Puppet module to manage your infrastructure, but you may want to download the module in order to inspect the code.

Download

Documentation

ptomulik/repoutil — version 0.1.5 Jun 14th 2016

ptomulik-repoutil

Build Status Coverage Status

Table of Contents

  1. Overview
  2. Module Description
  3. Setup
  4. Usage
  5. Reference
  6. Limitations
  7. Development

Overview

Puppet utilities to interact with package repositories. Simplify tasks such as obtaining lists of packages available for installation, their versions, installation candidates and so on. This plugin may be handy, if you need to implement facts describing packages available to agents' via their package repositories.

Module Description

The module provides means to access meta-information from package repositories such as apt or yum. The supported operations currently include:

  • listing available packages,
  • listing available package(s) versions,
  • retrieving full package records (containing descriptions, etc.),
  • determining installation candidate (version) for a given package,

The module currently supports the following providers:

  • apt, aptitude (Debian),
  • ports (FreeBSD)

Utilities provided by repoutil are devoted to developers of other modules, and are not intended to be used directly from puppet manifests.

Setup

What repoutil affects

  • it executes CLI commands necessary to query information from package repositories. The exact list of commands being executed depend on agent's OS and module's usage, but generally they should be regarded as harmless.

    The following CLI commands are currently used by repoutil's:

    • on Debian: apt-cache show|policy, aptitude show,
    • on FreeBSD, OpenBSD, NetBSD: make -C /path/to/ports search,

Setup Requirements

You may need to enable pluginsync in your puppet.conf.

Beginning with repoutil

Let's say, you're developing your custom fact or resource type and you need to characterize some packages existing in your package repository. The repoutil plugin may be used to obtain meta-data describing these packages.

The preliminary step to use repoutil utilities is to include its module file:

require 'puppet/util/repoutil'

Once the module file is included you may obtain reference to repoutil provider (call it repo), which implements access to a particular repository type. For example, apt repo may be obtained as follows:

repo = Puppet::Util.repoutil(:apt)

The :apt utility is suitable for use on Debian or Ubuntu. For other systems we should choose other facility. The most universal way is to load default utility for current environment:

repo = Puppet::Util.defaultrepoutil

Once we have obtained repo, we may request information from that repository. For example, we may wish to list all packages with names starting with 'apache':

apaches = repo.packages_with_prefix('apache')

This would return an array of all available package names starting with 'apache' prefix, for example:

["apache2-utils", "apache2-mpm-prefork", "apache2-mpm-worker", ... ]

There are also other methods. For example:

apache_versions = repo.package_versions_with_prefix('apache')

would return a hash with keys being the package names (from previous example) and arrays of available versions as values:

{
  "apache2"             => ["2.4.6-2", "2.2.22-13"],
  "apache2-mpm-prefork" => ["2.4.6-2", "2.2.22-13"],
  "apache2-mpm-itk"     => ["2.4.6-2", "2.2.22-13"],
  . . .
}

To retrieve full records from package database (including version, description, and other information) you may use:

apache_records = repo.package_records_with_prefix('apache')

This would return a hash such as the following:

{
  "apache2" => {
    "2.4.6-2" => {
      "Package"=>"apache2",
      "Version"=>"2.4.6-2",
      "Installed-Size"=>"481",
      . . .
    },
    "2.2.22-13" => {
      "Package"=>"apache2",
      "Version"=>"2.2.22-13",
      "Installed-Size"=>"29",
      . . .
    }
  },

  "apache2-mpm-prefork => {
    . . .
  },

  .
  .
  .
}

Of course, we may operate on exact package names, that is we may request information for one particular package. For example, in order to obtain available package versions we call:

apache2_versions = repo.package_versions('apache2')

This method returns an array, such as ["2.4.6-2", "2.2.22-13"] or nil (if there is no database entry for the package).

To see the installation candidate for package (what version of the package would be installed if we requested package installation/upgrade), we do:

apache2_candidates = repo.package_candidate('apache2')

This returns a string (e.g. "2.4.6-2") or nil (if there is no such package in repository).

To retrieve full records for available package versions we type:

apache2_records = repo.package_records('apache2')

This should return a hash such as:

{
  "2.4.6-2" =>
  {
    "Package"=>"apache2",
    "Version"=>"2.4.6-2",
    "Installed-Size"=>"481",
    . . .
  },
  "2.2.22-13" =>
  {
    "Package"=>"apache2",
    "Version"=>"2.2.22-13",
    "Installed-Size"=>"29",
    . . .
  }
}

In case there is no such package in repository, nil is returned.

Usage

The module consists of two ruby classes:

and few shorthand module-level methods.

The Puppet::Util::RepoUtil class abstracts CLI commands used to access repository cache/database and is inherited by several sub-classes implementing particular types of repositories (providers). The user actually operates on providers, that is on subclasses of Puppet::Util::RepoUtil. Each repoutil provider shall correspond to an appropriate package provider from puppet core (note, not all puppet providers are covered here). For example, there is Puppet::Util::RepoUtils::Apt class (:apt repoutil) which corresponds to :apt package provider. To retrieve appropriate repoutil, you should use one of the methods from Puppet::Util::RepoUtils described in reference. Methods within Puppet::Util::RepoUtil (and descendants) may be used to operate on a single repository.

Methods within Puppet::Util::RepoUtils are twofold. Some of them are provided to manage providers. These include newrepoutil (to implement new providers), unrepoutil (to unregister particular provider), repoutils (to retrieve all available providers), suitablerepoutils (to retrieve all the providers suitable for the current environment), defaultrepoutil to get provider default to current environment and repoutil (to retrieve particular provider). Other methods within Puppet::Util::RepoUtils may be used to perform collective operations on repositories. For example, package_candidates may be used to retrieve lists of package candidates known to all the suitable package repositories (this yields a hash of the form {:apt => {...}, :aptitude => {...}, ...}).

Reference

Methods within Puppet::Util module

newrepoutil(name, options = {}, &block)

Shorthand to Puppet::Util::RepoUtils.newrepoutil.

repoutil(name)

Shorthand to Puppet::Util::RepoUtils.repoutil.

repoutils

Shorthand to Puppet::Util::RepoUtils.repoutils.

suitablerepoutils

Shorthand to Puppet::Util::RepoUtils.suitablerepoutils.

defaultrepoutil

Shorthand to Puppet::Util::RepoUtils.defaultrepoutil.

Methods within Puppet::Util::RepoUtils class

Provider management

newrepoutil(name, options = {}, &block)

Define new repo provider. This is intended for developers/contributors and may be used to add support for new repository types. For details about extending repoutil module see adding new utility.

unrepoutil(name)

Unload/unregister a repoutil provider created with newrepoutil(name).

repoutil(name)

Retrieve repoutil provider identified by name. The provider must be first created with newrepoutil(name, ...).

Example:

repo = Puppet::Util::RepoUtils.repoutil(:apt)
candidate = repo.package_candidate('apache2')
repoutils

Retrieve all existing repoutil providers (includes also those not suitable for a given environment).

Example:

repos = Puppet::Util::RepoUtils.repoutils
apache2_candidates = {}
repos.each do |repo|
  if repo.suitable?
    apache2_candidates[repo] = repo.package_candidate('apache2')
  end
end
suitablerepoutils

Retrieve all repoutil providers suitable for the current environment.

Example:

repos = Puppet::Util::RepoUtils.suitablerepoutils
apache2_candidates = {}
repos.each do |repo|
  apache2_candidates[repo] = repo.package_candidate('apache2')
end
defaultrepoutil

Retrieve repoutil provider that is default to current environment.

Example:

repo = Puppet::Util::RepoUtils.defaultrepoutil
candidates = repo.package_candidate('apache2')
loadall

For internal use.

repoutilloader

For internal use.

Collective operations on repositories

package_records(packages, utils = suitablerepoutils)

Return package records obtained from multiple sources. This function performs query on multiple repositories at once. It may be used to gather package information from multiple repository types available to the local system.

Arguments:

  • packages - package name (exact) or an array of (exact) package names,
  • utils - array with repoutil providers to be queried (optional).

Example:

records = Puppet::Util::RepoUtils.package_records('apache2')

Would return a hash of the form:

{
  :aptitude =>
  {
    'apache2' =>
    {
      '2.2.13-2' => { 'Package' => 'apache2', 'Version' => '2.2.13-2', ... }
      '2.4.6-2'  => { ... }
    }
  },
  :apt  =>
  {
    'apache2' =>
    {
      '2.2.13-2' => { ... },
      '2.4.6-2'  => { ... }
    }
  },
   ...
 }

where :apt, :aptitude, ..., are keys corresponding to repoutil providers listed in utils.

package_versions(packages, utils = suitablerepoutils)

Return package versions available from multiple sources. This function performs query on multiple repositories at once. It may be used to gather package information from multiple repository types available to the local system.

Arguments:

  • packages - package name (exact) or an array of (exact) package names,
  • utils - array with repoutil providers to be queried (optional).

Example:

versions = Puppet::Util::RepoUtils.package_versions(['apache2', 'apache2-dev'])

Would return a hash of the form:

{
  :aptitude =>
  {
    "apache2"     =>  ["2.4.6-2"],
    "apache2-dev" =>  ["2.4.6-2"]
  },
  :apt =>
  {
    "apache2"     =>  ["2.4.6-2", "2.2.22-13"],
    "apache2-dev" =>  ["2.4.6-2"]
  }
}

where :apt, :aptitude, ..., are keys corresponding to repoutil providers listed in utils.

package_candidates(packages, utils = suitablerepoutils)

Return package candidates available for installation from multiple sources. This function may perform query on multiple repositories at once. It may be used to gather package information from multiple repository types available to the local system.

Arguments:

  • packages - package name (exact) or an array of (exact) package names,
  • utils - array with repoutil providers to be queried (optional).

Example:

candidates = Puppet::Util::RepoUtils.package_candidates(['apache2', 'apache2-dev'])

would return a hash of the form:

{
  :aptitude => { "apache2" => "2.4.6-2", "apache2-dev" => "2.4.6-2" },
  :apt      => { "apache2" => "2.4.6-2", "apache2-dev" => "2.4.6-2" }
}

where :apt, :aptitude, ..., are keys corresponding to repoutil providers listed in utils.

packages_with_prefixes(prefixes, utils = suitablerepoutils)

Search multiple repositories for packages having names starting with (one of the) prefixes. This function may perform search on multiple repositories at once. It may be used to gather package information from multiple repository types available to the local system.

Arguments:

  • prefixes - prefix or an array of prefixes to be matched against package names,
  • utils - array with repoutil providers to be queried (optional).

Example:

names = Puppet::Util::RepoUtils.packages_with_prefixes(['postfix', 'apache2'])

would return a hash of the form:

{
  :apt      => ["postfix-policyd-spf-python", "postfix-cdb", ..., "apache2", "apache2-data", ... ],
  :aptitude => ["postfix-policyd-spf-python", "postfix-cdb", ..., "apache2", "apache2-data", ... ]
}
package_records_with_prefixes(prefixes, utils = suitablerepoutils)

Search multiple repositories for packages having names starting with (one of the) prefixes and return their full records. This function may perform search on multiple repositories at once. It may be used to gather package information from multiple repository types available to the local system.

Arguments:

  • prefixes - prefix or an array of prefixes to be matched against package names,
  • utils - array with repoutil providers to be queried (optional).

Returns a hash with the form same as in Puppet::Util::RepoUtils.package_records

package_versions_with_prefixes(prefixes, utils = suitablerepoutils)

Search multiple repositories for packages having names starting with (one of the) prefixes and return their available versions. This function may perform search on multiple repositories at once. It may be used to gather package information from multiple repository types available to the local system.

Arguments:

  • prefixes - prefix or an array of prefixes to be matched against package names,
  • utils - array with repoutil providers to be queried (optional).

Returns a hash with the form same as in Puppet::Util::RepoUtils.package_versions

package_candidates_with_prefixes(prefixes, utils = suitablerepoutils)

Search multiple repositories for packages having names starting with (one of the) prefixes and return their installation candidates. This function may perform search on multiple repositories at once. It may be used to gather package information from multiple repository types available to the local system.

Arguments:

  • prefixes - prefix or an array of prefixes to be matched against package names,
  • utils - array with repoutil providers to be queried (optional).

Returns a hash with the form same as in Puppet::Util::RepoUtils.package_candidates

Methods within Puppet::Util::RepoUtil class

package_name_regexp

Regular expression to match package names. Used for validation of input arguments (validate_package_name) by functions having package argument and to parse output from CLI commands.

package_prefix_regexp

Regular expression to match package prefixes. Used for validation of input arguments (validate_package_prefix) by functions having prefix argument.

validate_package_name(package)

Throw an ArgumentError exception if package doesn't match the package_name_regexep.

validate_package_prefix(prefix)

Throw an ArgumentError exception if prefix doesn't match the package_prefix_regexep.

package_name_to_pattern(package)

Convert package name to a pattern for use by retrieve_records or retrieve_candidates.

package_prefix_to_pattern(package)

Convert package prefix to a pattern for use by retrieve_records or retrieve_candidates.

candidates_cache

Hash containing package candidates already seen by provider. The format is

{
  'package1' => 'ver1',
  'package2' => 'ver1',
  ...
}

Certain methods operating on exact package names, such as package_candidate, are searching this cache first and launchning external CLI tools only if the record is not found in this cache.

records_cache

Hash containing package records already seen by provider. The format is

{
  'package1' =>
  {
    'ver1' => { 'Field1' => 'Text1', 'Field2' => 'Text2', ... },
    'ver2' => { 'Field1' => 'Text1', 'Field2' => 'Text2', ... },
    ...
  },
  'package2' =>
  {
    'ver1' => { 'Field1' => 'Text1', 'Field2' => 'Text2', ... },
    'ver2' => { 'Field1' => 'Text1', 'Field2' => 'Text2', ... },
    ...
  },
  ...
}

Certain methods operating on exact package names, such as package_record are searching this cache first and launching external CLI tools only if the record is not found in the cache.

clear_candidates_cache

Clear candidates_cache.

clear_records_cache

Clear records_cache.

retrieve_candidates(pattern)

Retrieve information about package candidates from external source (CLI command). This method shall also update candidates_cache and may also update the records_cache. The pattern argument is a string containig tool-specific pattern used to search repository database - for example in :apt provider this is a pattern for apt-cache policy command.

retrieve_records(pattern)

Retrieve package records from external source (CLI command). This method shall also update records_cache and may also update the candidates_cache. The pattern argument is a string containing tool-specific pattern used to search repository database - for example in :apt provider this is a pattern for apt-cache show command.

package_records(package)

Return package records describing given package. The package argument shall be an exact package name. The returned hash has the following form

{
  'ver1' => { 'Field1' => 'Text1', 'Field2' => 'Text2' },
  'ver2' => { 'Field1' => 'Text1', 'Field2' => 'Text2' },
  ...
}

If package is not found in repository, nil is returned.

package_versions(package)

Return package versions available for installation. The package argument shall be an exact package name. The method returns an array containing available versions.

[ 'ver1', 'ver2', ... ]

If package is not found in repository, nil is returned.

package_candidate(package)

Return package's installation candidate. The package argument shall be an exact package name. The method returns a string containing version number of the installation candidate. If package is not found in repository, nil is returned.

packages_with_prefix(prefix)

List package names for all available packages having name starting with prefix.

package_versions_with_prefix(prefix)

Return package versions for all available packages having name starting with prefix. The function returns a hash in the form

{
  'packageA' => ['verA1', 'verA2', ...],
  'packageB' => ['verB1', 'verB2', ...],
}

If there is no package matching the prefix, an empty hash is returned.

package_candidates_with_prefix(prefix)

Return package candidates for all available packages having name starting with prefix. The function returns a hash in the form

{
  'package1' => 'ver1',
  'package2' => 'ver2',
}
package_records_with_prefix(prefix)

Return full records for all available packages having name starting with prefix. The function returns a hash in the form

{
  'packageA' =>
  {
    'verA1' => { 'Field1' => 'Text1', 'Field2' => 'Text2', ... },
    'verA2' => { 'Field1' => 'Text1', 'Field2' => 'Text2', ... },
    ...
  },
  'packageB' =>
  {
    'verB1' => { 'Field1' => 'Text1', 'Field2' => 'Text2', ... },
    'verB2' => { 'Field1' => 'Text1', 'Field2' => 'Text2', ... },
    ...
  },
  ...
}

If there is no package matching the prefix, an empty hash is returned.

Limitations

  • Currently supports only Debian/Ubuntu apt, aptitude and FreeBSD ports
  • Some tests are missing.

Development

The project is held on github:

https://github.com/ptomulik/puppet-repoutil

Feel free to submit bug reports, feature requests or to create pull requests.

Extending repoutil

Adding new utility (provider)

The new utility may be defined by using Puppet::Util.newrepoutil method.

  • Puppet::Util.newrepoutil(name, options = {}, &block)

If you're extending repoutil module, the source code of the new provider (say foo) should go to file

  • lib/puppet/util/repoutil/foo.rb

You may also consider adding :foo entry to utils array in

  • spec/unit/puppet/util/repoutil_spec.rb

to enable common unit tests for your provider. Specific tests should be implemented in

  • spec/unit/puppet/util/repoutil/foo_spec.rb

When defining new utility, one should define following methods in the block provided to newrepoutil:

  • self.package_name_regexp
  • self.package_prefix_regexp
  • self.package_name_to_pattern(package)
  • self.package_prefix_to_pattern(prefix)
  • self.retrieve_candidates(pattern)
  • self.retrieve_records(pattern)

To configure suitability, confines, defaults etc., use same methods as in normal providers.

Example template for a repoutil foo is presented below:

# lib/puppet/util/repoutil/foo.rb
module Puppet::Util
  newrepoutil(:foo) do
    commands :foocmd => '/usr/bin/foo'

    def self.package_name_regexp
      # this is tool-specific, check if this pattern fits your needs
      /[a-z0-9][a-z0-9\.+-]+/
    end

    def self.package_prefix_regexp
      # this is tool-specific, check if this pattern fits your needs
      /(?:[a-z0-9][a-z0-9\.+-]*)?/
    end

    # escape characters that could be interpreted as meta-characters
    # by CLI commands used for repository lookups
    def self.escape_package_name(package)
      # this is tool-specific, check if this pattern fits your needs
      package.gsub(/([\.\+])/) {|c| '\\' + c}
    end

    # escape characters that could be interpreted as meta-characters
    # by CLI commands used for repository lookups
    def self.escape_package_prefix(prefix)
      # this is tool-specific, check if this pattern fits your needs
      prefix.gsub(/([\.\+])/) {|c| '\\' + c}
    end

    def self.package_name_to_pattern(package)
      # this is tool-specific, check if this conversion fits your needs
      "^#{escape_package_name(package)}$"
    end

    def self.package_prefix_to_pattern(prefix)
      # this is tool-specific, check if this conversion fits your needs
      "^#{escape_package_prefix(prefix)}"
    end

    def self.show_candidates(pattern)
      # adjust to the actual syntax of your CLI command
      foocmd 'candidates',  pattern
    end

    def self.show_records(pattern)
      # adjust to the actual syntax of your CLI command
      foocmd 'records', pattern
    end

    def self.retrieve_candidates(pattern)
      output = show_candidates(pattern).chomp
      # now, extract candidates from output, update candidates_cache and
      # return the extracted candidates
    end

    def self.retrieve_records(pattern)
      output = show_records(pattern).chomp
      # now, extract records from output, update records_cache and
      # return the extracted records
    end
  end
end