pluginlib 0.9.1

Features

• Plugins are validated when they are imported

• Plugins can be loaded through different mechanisms (modules, filesystem paths, entry points)

• Multiple versions of the same plugin are supported (The newest one is used by default)

• Plugins can be blacklisted by type, name, or version

• Multiple plugin groups are supported so one program can use multiple sets of plugins that won’t conflict

• Plugins support conditional loading (examples: os, version, installed software, etc)

• Once loaded, plugins can be accessed through dictionary or dot notation

Fedora and EL (RHEL/CentOS/Rocky/Alma)

(EPEL repositories must be configured for EL8)

$ dnf install python3-pluginlib

EL7 (RHEL/CentOS)

(EPEL repositories must be configured)

$ yum install python2-pluginlib
$ yum install python36-pluginlib

Step 1: Define plugin parent classes

All plugins are subclasses of parent classes. To create a parent class, use the @Parent decorator.

The @Parent decorator can take a plugin type for accessing child plugins of the parent. If a plugin type isn’t given, the class name will be used.

The @Parent decorator can also take a group keyword which restricts plugins to a specific plugin group. group should be specified if plugins for different projects could be accessed in an single program, such as with libraries and frameworks. For more information, see the Plugin Groups section.

Methods required in child plugins should be labeled as abstract methods. Plugins without these methods or with parameters that don’t match, will not be loaded. For more information, see the Abstract Methods section.

"""
sample.py
"""
import pluginlib

@pluginlib.Parent('parser')
class Parser(object):

    @pluginlib.abstractmethod
    def parse(self, string):
        pass

Step 2: Define plugin classes

To create a plugin, subclass a parent class and include any required methods.

Plugins can be customized through optional class attributes:

_alias_

Changes the name of the plugin which defaults to the class name.

_version_

Sets the version of the plugin. Defaults to the module __version__ or None If multiple plugins with the same type and name are loaded, the plugin with the highest version is used. For more information, see the Versions section.

_skipload_

Specifies the plugin should not be loaded. This is useful when a plugin is a parent class for additional plugins or when a plugin should only be loaded under certain conditions. For more information see the Conditional Loading section.

"""
sample_plugins.py
"""
import json
import sample

class JSON(sample.Parser):

    _alias_ = 'json'

    def parse(self, string):
        return json.loads(string)

Step 3: Load plugins

Plugins are loaded when the module they are in is imported. PluginLoader will load modules from specified locations and provides access to them.

PluginLoader can load plugins from several locations.

• A program’s standard library

• Entry points

• A list of modules

• A list of filesystem paths

Plugins can also be filtered through blacklists and type filters. See the Blacklists and Type Filters sections for more information.

Plugins are accessible through the PluginLoader.plugins property, a nested dictionary accessible through dot notation. For other ways to access plugins, see the Accessing Plugins section.

import pluginlib
import sample

loader = pluginlib.PluginLoader(modules=['sample_plugins'])
plugins = loader.plugins
parser = plugins.parser.json()
print(parser.parse('{"json": "test"}'))