Red Hat Training

A Red Hat training course is available for Red Hat Enterprise Linux

5. Developing Installer Add-ons

5.1. Introduction to Anaconda and Add-ons

5.1.1. Introduction to Anaconda

Anaconda is the operating system installer used in Fedora, Red Hat Enterprise Linux, and their derivatives. It is a set of Python modules and scripts together with some additional files like Gtk widgets (written in C), systemd units, and dracut libraries. Together, they form a tool that allows users to set parameters of the resulting (target) system and then set such a system up on a machine. The installation process has four major steps:
  • installation destination preparation (usually disk partitioning)
  • package and data installation
  • boot loader installation and configuration
  • configuration of the newly installed system
There are three ways you can control the installer and specify installation options. The most common approach is to use the graphical user interface (GUI). This interface is meant to allow users to install the system interactively with little or no configuration required before beginning the installation, and it should cover all common use cases, including setting up complicated partitioning layouts.
The graphical interface also supports remote access over VNC, which allows you to use the GUI even on systems with no graphics cards or even attached monitor. However, there are still cases where this is not desired, but at the same time, you may want to perform an interactive installation. For these cases, a text mode (TUI) is available. The TUI works in a way similar to a monochrome line printer, which allows it to work even on serial consoles which do not support cursor movement, colors and other advanced features. The text mode is limited in that it only allows you to customize most common options, such as network settings, language options or installation (package) source; advanced features such as manual partitioning are not available in this interface.
The third way to install a system using Anaconda is by using a Kickstart file - a plain text file with shell-like syntax which can contain data to drive the installation process. A Kickstart file allows you to partially or completely automate the installation. A certain set of commands which configures all required areas is necessary to completely automate the installation; if one or more of the required commands is missing, the installation will require interaction. If all required commands are present, the installation will be performed in a completely automatic way, without any need for interaction.
Kickstart provides the highest amount of options, covering use cases where neither the TUI nor the GUI is sufficient. Every feature in Anaconda must always be supported in Kickstart; other interfaces follow only subsets of all available options, which allows them to remain clear.

5.1.2. Firstboot and Initial Setup

The first boot of the newly installed system is traditionally considered a part of the installation process as well, because some parts of configuration such as user creation are often performed at this point. Previously, the Firstboot tool has been used for this purpose, allowing you to register your newly installer Red Hat Enterprise Linux system or configure Kdump. However, Firstboot relies on no longer maintained tools such as Gtk2 and the pygtk2 module. [1] For this reason, a new tool called Initial Setup was developed, which reuses code from Anaconda. This allows add-ons developed for Anaconda to be easily reused in Initial Setup. This topic is further discussed in Section 5.6, “Writing an Anaconda add-on”.

5.1.3. Anaconda and Initial Setup Add-ons

Installing a new operating system is a vastly complicated use case - each user may want to do something slightly different. Designing an installer for every corner case would cause it to be cluttered with rarely-used functionality. For this reason, when the installer was being rewritten into its current form, it gained support for add-ons.
Anaconda add-ons can be used to add your own Kickstart commands and options as well as new configuration screens in the graphical and text-based user interface, depending on your specific use case. Each add-on must have Kickstart support; the GUI and TUI are optional, but can be very helpful.
In current releases of Red Hat Enterprise Linux (7.1 and later) and Fedora [2] (21 and later), one add-on is included by default: The Kdump add-on, which adds support for configuring kernel crash dumping during the installation. This add-on has full support in Kickstart (using the %addon com_redhat_kdump command and its options) and is fully integrated as an additional screen in the text-based and graphical interfaces. You can develop other add-ons in the same way and add them to the default installer using procedures described further in this guide.

5.1.4. Additional Information

Following links contain additional information about Anaconda and Initial Setup:

5.2. Architecture of Anaconda

Anaconda is a set of Python modules and scripts. It also uses several external packages and libraries, some of which were created specifically for the installer. Major components of this toolset include the following packages:
  • pykickstart - used to parse and validate Kickstart files and also to provide a data structure which stores values which drive the installation
  • yum - the package manager which handles installation of packages and resolving dependencies
  • blivet - originally split from the anaconda package as pyanaconda.storage; used to handle all activities related to storage management
  • pyanaconda - package containing the core of the user interface and modules for functionality unique to Anaconda, such as keyboard and timezone selection, network configuration, and user creation, as well as a number of utilities and system-oriented functions
  • python-meh - contains an exception handler which gathers and stores additional system information in case of a crash and passes this information to the libreport library, which itself is a part of the ABRT Project.
The life cycle of data during the installation process is straightforward. If a Kickstart file is provided, it is processed by the pykickstart module and imported into memory as a tree-like structure. If no Kickstart file is provided, an empty tree-like structure is created instead. If the installation is interactive (not all required Kickstart commands have been used), the structure is then updated with choices made by the user in the interactive interface.
Once all required choices are made, the installation process begins and values stored in the structure are used to determine parameters of the installation. The values are also written as a Kickstart file which is saved in the /root/ directory on the installed system; therefore the installation can be replicated automatically by reusing this automatically generated Kickstart file.
Elements of the tree-like structure are defined by the pykickstart package, but some of them can be overriden by modified versions from the pyanaconda.kickstart module. An important rule which governs this behavior is that there is no place to store configuration data, and the installation process is data-driven and relies on transactions as much as possible. This enforces the following features:
  • every feature of the installer must be supported in Kickstart
  • there is a single, obvious point in the installation process where changes are written to the target system; before this point, no lasting changes (e.g. formatting storage) are made
  • every change made manually in the user interface is reflected in the resulting Kickstart file and can be replicated
The fact that the installation is data-driven means that installation and configuration logic lies within the methods of the items in the tree-like structure. Every item is set up (the setup method) to modify the runtime environment of the installation if necessary, and then executed (the execute method) to perform the changes on the target system. These methods are further described in Section 5.6, “Writing an Anaconda add-on”.

5.3. The Hub & Spoke model

One of the notable differences between Anaconda and most other operating system installers is its non-linear nature, also known as the hub and spoke model.
The hub and spoke model of Anaconda has several advantages, including:
  • users are not forced to go through the screens in some strictly defined order
  • users are not forced to visit every screen no matter if they understand what the options configured in it mean or not
  • it is good for the transactional mode where all desired values can be set while nothing is actually happening to the underlying machine until a special button is clicked
  • it provides way to show an overview of the configured values
  • it has a great support for extensibility, because additional spokes can be put on hubs without need to reorder anything and resolve some complex ordering dependencies
  • it can be used for both graphical and text mode of the installer
The diagram below shows the installer layout as well as possible interactions between hubs and spokes (screens):

Figure 2. Diagram of the hub and spoke model

In the diagram, screens 2-13 are called normal spokes, and screens 1 and 14 are standalone spokes. Standalone spokes are a type of screen which is a type of screen that should be used only in case it has to be visited before (or after) the following (or previous) standalone spoke or hub. This may be, for example, the Welcome screen at the beginning of the installation which prompts you to choose your language for the rest of the installation.

Note

Screens mentioned in the rest of this section are screens from the installer's graphical interface (GUI).
Central points of the hub and spoke model are hubs. There are two hubs by default:
  • The Installation Summary hub which shows a summary of configured options before the installation begins
  • The Configuration and Progress hub which appears after you click Begin Installation in Installation Summary, and which displays the progress of the installation process and allows you to configure additional options (set the root password and create a user account).
Each spoke has several predefined properties which are reflected on the hub. These are:
  • ready - states whether the spoke can be visited or not; for example, when the installer is configuring a package source, that spoke is not ready, is colored gray, and cannot be accessed until configuration is complete
  • completed - marks the spoke as completed (all required values are set) or not
  • mandatory - determines whether the spoke must be visited and confirmed by the user before continuing the installation; for example, the Installation Destination spoke must always be visited, even if you want to use automatic disk partitioning
  • status - provides a short summary of values configured within the spoke (displayed under the spoke name in the hub)
To make the user interface clearer, spokes are grouped together into categories. For example, the Localization category groups together spokes for keyboard layout selection, language support and time zone settings.
Each spoke contains UI controls which display and allow you to modify values from one or more sub-trees of the in-memory tree-like structure which was discussed in Section 5.2, “Architecture of Anaconda”. As Section 5.6, “Writing an Anaconda add-on” explains, the same applies to spokes provided by add-ons.

5.4. Threads and Communication

Some of the actions which need to be performed during the installation process, such as scanning disks for existing partitions or downloading package metadata, can take a long time. To prevent you from waiting and remain responsive if possible, Anaconda runs these actions in separate threads.
The Gtk toolkit does not support element changes from multiple threads. The main event loop of Gtk runs in the main thread of the Anaconda process itself, and all code performing actions which involve the GUI must make sure that these actions are run in the main thread as well. The only supported way to do so is by using the GLib.idle_add, which is not always easy or desired. To alleviate this problem, several helper functions and decorators are defined in the pyanaconda.ui.gui.utils module.
The most useful of those are the @gtk_action_wait and @gtk_action_nowait decorators. They change the decorated function or method in such a way that when this function or method is called, it is automatically queued into Gtk's main loop, run in the main thread, and the return value is either returned to the caller or dropped, respectively.
As mentioned previously, one of the main reasons for using multiple threads is to allow the user to configure some screens while other screens which are currently busy (such as Installation Source when it downloads package metadata) configure themselves. Once the configuration is finished, the spoke which was previously busy needs to announce that it is now ready and not blocked; this is handled by a message queue called hubQ, which is being periodically checked in the main event loop. When a spoke becomes accessible, it sends a message to this queue announcing this change and that it should no longer be blocked.
The same applies in a situation where a spoke needs to refresh its status or completion flag. The Configuration and Progress hub has a different queue called progressQ which serves as a medium to transfer installation progress updates.
These mechanisms are also needed for the text-based interface, where the situation is more complicated; there is no main loop in text mode, instead the majority of time in this mode is spent waiting for keyboard input.

5.5. Anaconda Add-on Structure

An Anaconda add-on is a Python package containing a directory with an __init__.py and other source directories (subpackages) inside. Because Python allows importing each package name only once, the package top-level directory name must be unique. At the same time, the name can be arbitrary, because add-ons are loaded regardless of their name - the only requirement is that they must be placed in a specific directory.
The suggested naming convention for add-ons is therefore similar to Java packages or D-Bus service names: prefix the add-on name with the reversed domain name of your organization, using underscores (_) instead of dots so that the directory name is a valid identifier for a Python package. An example add-on name following these suggestions would therefore be e.g. com_example_hello_world. This convention follows the recommended naming scheme for Python package and module names.

Important

Make sure to create an __init__.py file in each directory. Directories missing this file are not considered valid Python packages.
When writing an add-on, keep in mind that every function supported in the installer must be supported in Kickstart; GUI and TUI support is optional. Support for each interface (Kickstart, graphical interface and text interface) must be in a separate subpackage and these subpackages must be named ks for Kickstart, gui for the graphical interface and tui for the text-based interface. The gui and tui packages must also contain a spokes subpackage. [3]
Names of modules inside these packages are arbitrary; the ks/, gui/ and tui/ directories can contain Python modules with any name.
A sample directory structure for an add-on which supports every interface (Kickstart, GUI and TUI) will look similar to the following:

Example 2. Sample Add-on Structure

com_example_hello_world
├─ ks
│  └─ __init__.py
├─ gui
│  ├─ __init__.py
│  └─ spokes
│     └─ __init__.py
└─ tui
   ├─ __init__.py
   └─ spokes
   └─ __init__.py
Each package must contain at least one module with an arbitrary name defining classes inherited from one or more classes defined in the API. This is further discussed in Section 5.6, “Writing an Anaconda add-on”.
All add-ons should follow Python's PEP 8 and PEP 257 guidelines for docstring conventions. There is no consensus on the format of the actual content of docstrings in Anaconda; the only requirement is that they are human-readable. If you plan to use automatically generated documentation for your add-on, docstrings should follow the guidelines for the toolkit you use to accomplish this.

5.6. Writing an Anaconda add-on

The sections below will demonstrate the process writing and testing a sample add-on called Hello World. This sample add-on will support all interfaces (Kickstart, GUI and TUI). Sources for this sample add-on are available on GitHub in the rhinstaller/hello-world-anaconda-addon repository; it is recommended to clone this repository or at least open the sources in the web interface.
Another repository to review is rhinstaller/anaconda, which contains the installer source code; it will be referred to in several parts of this section as well.
Before you begin developing the add-on itself, start by creating its directory structure as described in Section 5.5, “Anaconda Add-on Structure”. Then, continue with Section 5.6.1, “Kickstart Support”, as Kickstart support is mandatory for all add-ons. After that, you can optionally continue with Section 5.6.2, “Graphical user interface” and Section 5.6.3, “Text User Interface” if needed.

5.6.1. Kickstart Support

Kickstart support is always the first part of any add-on that should be developed. Other packages - support for the graphical and text-based interface - will depend on it. To begin, navigate to the com_example_hello_world/ks/ directory you have created previously, make sure it contains an __init__.py file, and add another Python script named hello_world.py.
Unlike built-in Kickstart commands, add-ons are used in their own sections. Each use of an add-on in a Kickstart file begins with an %addon statement and is closed by %end. The %addon line also contains the name of the add-on (such as %addon com_example_hello_world) and optionally a list of arguments, if the add-on supports them.
An example use of an add-on in a Kickstart file looks like the example below:

Example 3. Using an Add-on in a Kickstart File

%addon ADDON_NAME [arguments]
first line
second line
...
%end
The key class for Kickstart support in add-ons is called AddonData. This class is defined in pyanaconda.addons and represents an object for parsing and storing data from a Kickstart file.
Arguments are passed as a list to an instance of the add-on class inherited from the AddonData class. Anything between the first and last line is passed to the add-on's class one line at a time. To keep the example Hello World add-on simple, it will merge all lines in this block into a single line and separate the original lines with a space.
The example add-on requires a class inherited from AddonData with a method for handling the list of arguments from the %addon line, and a method for handling lines inside the section. The pyanaconda/addons.py module contains two methods which can be used for this:
  • handle_header - takes a list of arguments from the %addon line (and line numbers for error reporting)
  • handle_line - takes a single line of content from between the %addon and %end statements
The example below demonstrates a Hello World add-on which uses the methods described above:

Example 4. Using handle_header and handle_line

from pyanaconda.addons import AddonData
from pykickstart.options import KSOptionParser

# export HelloWorldData class to prevent Anaconda's collect method from taking
# AddonData class instead of the HelloWorldData class
# :see: pyanaconda.kickstart.AnacondaKSHandler.__init__
__all__ = ["HelloWorldData"]

HELLO_FILE_PATH = "/root/hello_world_addon_output.txt"

class HelloWorldData(AddonData):
    """
    Class parsing and storing data for the Hello world addon.

    :see: pyanaconda.addons.AddonData

    """

def __init__(self, name):
    """
    :param name: name of the addon
    :type name: str

    """

    AddonData.__init__(self, name)
    self.text = ""
    self.reverse = False

def handle_header(self, lineno, args):
    """
    The handle_header method is called to parse additional arguments in the
    %addon section line.

    :param lineno: the current linenumber in the kickstart file
    :type lineno: int
    :param args: any additional arguments after %addon <name>
    :type args: list

    """

    op = KSOptionParser()
    op.add_option("--reverse", action="store_true", default=False,
    dest="reverse", help="Reverse the display of the addon text")
    (opts, extra) = op.parse_args(args=args, lineno=lineno)

    # Reject any additoinal arguments. Since AddonData.handle_header
    # rejects any arguments, we can use it to create an error message
    # and raise an exception.
    if extra:
        AddonData.handle_header(self, lineno, extra)

    # Store the result of the option parsing
    self.reverse = opts.reverse

def handle_line(self, line):
    """
    The handle_line method that is called with every line from this addon's
    %addon section of the kickstart file.

    :param line: a single line from the %addon section
    :type line: str

    """

    # simple example, we just append lines to the text attribute
    if self.text is "":
        self.text = line.strip()
    else:
        self.text += " " + line.strip()
The example begins by importing necessary methods and defining an __all__ variable which is necessary to prevent Anaconda's collect method from taking the AddonData class instead of add-on specific HelloWorldData.
Then, the example shows a definition of the HelloWorldData class inherited from AddonData with its __init__ method calling the parent's __init__ and initializing the attributes self.text and self.reverse to False.
The self.reverse attribute is populated in the handle_header method, and the self.text is populated in handle_line. The handle_header method uses an instance of the KSOptionParser provided by pykickstart to parse additional options used on the %addon line, and handle_line strips the content lines of white space at the beginning and end of each line, and appends them to self.text.
The code above covers the first phase of the data life cycle in the installation process: it reads data from the Kickstart file. The next step is to use this data to drive the installation process. Two predefined methods are available for this purpose:
  • setup - called before the installation transaction starts and used to make changes to the installation runtime environment
  • execute - called at the end of the transaction and used to make changes to the target system
To use these two methods, you must add some new imports and a constant to your module, as shown in the following example:

Example 5. Importing the setup and execute Methods

import os.path

from pyanaconda.addons import AddonData
from pyanaconda.constants import ROOT_PATH

HELLO_FILE_PATH = "/root/hello_world_addon_output.txt"
An updated example of the Hello World add-ons with the setup and execute methods included is below:

Example 6. Using the setup and execute Methods

def setup(self, storage, ksdata, instclass, payload):
    """
    The setup method that should make changes to the runtime environment
    according to the data stored in this object.

    :param storage: object storing storage-related information
    (disks, partitioning, bootloader, etc.)
    :type storage: blivet.Blivet instance
    :param ksdata: data parsed from the kickstart file and set in the
    installation process
    :type ksdata: pykickstart.base.BaseHandler instance
    :param instclass: distribution-specific information
    :type instclass: pyanaconda.installclass.BaseInstallClass
    :param payload: object managing packages and environment groups
    for the installation
    :type payload: any class inherited from the pyanaconda.packaging.Payload
    class
    """

    # no actions needed in this addon
    pass

def execute(self, storage, ksdata, instclass, users, payload):
    """
    The execute method that should make changes to the installed system. It
    is called only once in the post-install setup phase.

    :see: setup
    :param users: information about created users
    :type users: pyanaconda.users.Users instance

    """

    hello_file_path = os.path.normpath(ROOT_PATH + HELLO_FILE_PATH)
    with open(hello_file_path, "w") as fobj:
        fobj.write("%s\n" % self.text)
In the above example, the setup method does nothing; the Hello World add-on does not make any changes to the installation runtime environment. The execute method writes stored text into a file created in the target system's root (/) directory.
The most important information in the above example is the amount and meaning of the arguments passed to the two new methods; these are described in docstrings within the example.
The final phase of the data life cycle, as well as the last part of the code needed in a module providing Kickstart support, is generating a new Kickstart file, which includes values set at installation time, at the end of the installation process as described in Section 5.2, “Architecture of Anaconda”. This is performed by calling the __str__ method recursively on the tree-like structure storing installation data, which means that the class inherited from AddonData must define its own __str__ method which returns its stored data in valid Kickstart syntax. This returned data must be possible to parse again using pykickstart.
In the Hello World example, the __str__ method will be similar to the following example:

Example 7. Defining a __str__ Method

def __str__(self):
    """
    What should end up in the resulting kickstart file, i.e. the %addon
    section containing string representation of the stored data.

    """

    addon_str = "%%addon %s" % self.name

    if self.reverse:
        addon_str += "--reverse"

    addon_str += "\n%s\n%%end" % self.text
    return addon_str
Once your Kickstart support module contains all necessary methods (handle_header, handle_line, setup, execute and __str__), it becomes a valid Anaconda add-on. You can continue with the following sections to add support for the graphical and text-based user interfaces, or you can continue with Section 5.7, “Deploying and testing an Anaconda add-on” and test the add-on.

5.6.2. Graphical user interface

This section will describe adding support for the graphical user interface (GUI) to your add-on. Before you begin, make sure that your add-on already includes support for Kickstart as described in the previous section.

Note

Before you start developing add-ons with support for the graphical interface, make sure to install the anaconda-widgets and anaconda-widgets-devel packages, which contain Gtk widgets specific for Anaconda such as SpokeWindow.
5.6.2.1. Basic features
Similarly to Kickstart support in add-ons, GUI support requires every part of the add-on to contain at least one module with a definition of a class inherited from a particular class defined by the API. In case of graphical support, the only recommended class is NormalSpoke, which is defined in pyanaconda.ui.gui.spokes. As the class name suggests, it is a class for the normal spoke type of screen as described in Section 5.3, “The Hub & Spoke model”.
To implement a new class inherited from NormalSpoke, you must define the following class attributes which are required by the API:
  • builderObjects - lists all top-level objects from the spoke's .glade file that should be, with their children objects (recursively), exposed to the spoke - or should be an empty list if everything should be exposed to the spoke (not recommended)
  • mainWidgetName - contains the id of the main window widget [4] as defined in the .glade file
  • uiFile - contains the name of the .glade file
  • category - contains the class of the category the spoke belongs to
  • icon - contains the identifier of the icon that will be used for the spoke on the hub
  • title defines the title that will be used for the spoke on the hub
Example module with all required definitions is shown in the following example:

Example 8. Defining Attributes Required for the Normalspoke Class

# will never be translated
_ = lambda x: x
N_ = lambda x: x

# the path to addons is in sys.path so we can import things from org_fedora_hello_world
from org_fedora_hello_world.gui.categories.hello_world import HelloWorldCategory
from pyanaconda.ui.gui.spokes import NormalSpoke

# export only the spoke, no helper functions, classes or constants
__all__ = ["HelloWorldSpoke"]

class HelloWorldSpoke(NormalSpoke):
    """
    Class for the Hello world spoke. This spoke will be in the Hello world
    category and thus on the Summary hub. It is a very simple example of
    a unit for the Anaconda's graphical user interface.

    :see: pyanaconda.ui.common.UIObject
    :see: pyanaconda.ui.common.Spoke
    :see: pyanaconda.ui.gui.GUIObject

    """

    ### class attributes defined by API ###

    # list all top-level objects from the .glade file that should be exposed
    # to the spoke or leave empty to extract everything
    builderObjects = ["helloWorldSpokeWindow", "buttonImage"]

    # the name of the main window widget
    mainWidgetName = "helloWorldSpokeWindow"

    # name of the .glade file in the same directory as this source
    uiFile = "hello_world.glade"

    # category this spoke belongs to
    category = HelloWorldCategory

    # spoke icon (will be displayed on the hub)
    # preferred are the -symbolic icons as these are used in Anaconda's spokes
    icon = "face-cool-symbolic"

    # title of the spoke (will be displayed on the hub)
    title = N_("_HELLO WORLD")
The __all__ attribute is used to export the spoke class, followed by the first lines of its definition including definitions of attributes mentioned above. The values of these attributes are referencing widgets defined in com_example_hello_world/gui/spokes/hello.glade file.
Two other notable attributes are present. The first is category, which has its value imported from the HelloWorldCategory class from the com_example_hello_world.gui.categories module. The HelloWorldCategory class will be discussed later, but for now, note that the path to add-ons is in sys.path so that things can be imported from the com_example_hello_world package.
The second notable attribute in the example is title, which contains two underscores in its definition. The first one is part of the N_ function name which marks the string for translation, but returns the non-translated version of the string (translation is done later). The second underscore marks the beginning of the title itself and makes the spoke reachable using the Alt+H keyboard shortcut.
What usually follows the header of the class definition and the class attributes definitions is the constructor that initializes an instance of the class. In case of the Anaconda graphical interface objects there are two methods initializing a new instance: the __init__ method and the initialize method.
The reason for two such functions is that the GUI objects may be created in memory at one time and fully initialized (which can take a longer time) at a different time. Therefore, the __init__ method should only call the parent's __init__ method and (for example) initialize non-GUI attributes. On the other hand, the initialize method that is called when the installer's graphical user interface initializes should finish the full initialization of the spoke.
In the sample Hello World add-on, these two methods are defined as follows (note the number and description of the arguments passed to the __init__ method):

Example 9. Defining the __init__ and initialize Methods

def __init__(self, data, storage, payload, instclass):
    """
    :see: pyanaconda.ui.common.Spoke.__init__
    :param data: data object passed to every spoke to load/store data
    from/to it
    :type data: pykickstart.base.BaseHandler
    :param storage: object storing storage-related information
    (disks, partitioning, bootloader, etc.)
    :type storage: blivet.Blivet
    :param payload: object storing packaging-related information
    :type payload: pyanaconda.packaging.Payload
    :param instclass: distribution-specific information
    :type instclass: pyanaconda.installclass.BaseInstallClass

    """

    NormalSpoke.__init__(self, data, storage, payload, instclass)

def initialize(self):
    """
    The initialize method that is called after the instance is created.
    The difference between __init__ and this method is that this may take
    a long time and thus could be called in a separated thread.

    :see: pyanaconda.ui.common.UIObject.initialize

    """

    NormalSpoke.initialize(self)
    self._entry = self.builder.get_object("textEntry")
Note the data parameter passed to the __init__ method. This is the in-memory tree-like representation of the Kickstart file where all data is stored. In one of the ancestors' __init__ methods it is stored in the self.data attribute, which allows all other methods in the class to read and modify the structure.
Because the HelloWorldData class has already been defined in Section 5.6.1, “Kickstart Support”, there already is a subtree in self.data for this add-on, and its root (an instance of the class) is available as self.data.addons.com_example_hello_world.
One of the other things an ancestor's __init__ does is initializing an instance of the GtkBuilder with the spoke's .glade file and storing it as self.builder. This is used in the initialize method to get the GtkTextEntry used to show and modify the text from the kickstart file's %addon section.
The __init__ and initialize methods are both important when the spoke is created. However, the main role of the spoke is to be visited by an user who wants to change or review the values this spoke shows and sets. To enable this, three other methods are available:
  • refresh - called when the spoke is about to be visited; This method refreshes the state of the spoke (mainly its UI elements) to make sure that current values stored in the self.data structure are displayed
  • apply - called when the spoke is left and used to store values from UI elements back into the self.data structure
  • execute - called when the spoke is left and used to perform any runtime changes based on the new state of the spoke
These functions are implemented in the sample Hello World add-on in the following way:

Example 10. Defining the refresh, apply and execute Methods

def refresh(self):
    """
    The refresh method that is called every time the spoke is displayed.
    It should update the UI elements according to the contents of
    self.data.

    :see: pyanaconda.ui.common.UIObject.refresh

    """

    self._entry.set_text(self.data.addons.org_fedora_hello_world.text)

def apply(self):
    """
    The apply method that is called when the spoke is left. It should
    update the contents of self.data with values set in the GUI elements.

    """

    self.data.addons.org_fedora_hello_world.text = self._entry.get_text()

def execute(self):
  """
  The excecute method that is called when the spoke is left. It is
  supposed to do all changes to the runtime environment according to
  the values set in the GUI elements.

  """

  # nothing to do here
  pass
You can use several additional methods to control the spoke's state:
  • ready - determines whether the spoke is ready to be visited; if the value is false, the spoke is not accessible (e.g. the Package Selection spoke before a package source is configured)
  • completed - determines if the spoke has been completed
  • mandatory - determines if the spoke is mandatory or not (e.g. the Installation Destination spoke, which must be always visited, even if you want to use automatic partitioning)
All of these attributes need to be dynamically determined based on the current state of the installation process. Below is a sample implementation of these methods in the Hello World add-on, which requires some value to be set in the text attribute of the HelloWorldData class:

Example 11. Defining the ready, completed and mandatory Methods

@property
def ready(self):
    """
    The ready property that tells whether the spoke is ready (can be visited)
    or not. The spoke is made (in)sensitive based on the returned value.

    :rtype: bool

    """

    # this spoke is always ready
    return True

@property
def completed(self):
    """
    The completed property that tells whether all mandatory items on the
    spoke are set, or not. The spoke will be marked on the hub as completed
    or uncompleted acording to the returned value.

    :rtype: bool

    """

    return bool(self.data.addons.org_fedora_hello_world.text)

@property
def mandatory(self):
    """
    The mandatory property that tells whether the spoke is mandatory to be
    completed to continue in the installation process.

    :rtype: bool

    """

    # this is an optional spoke that is not mandatory to be completed
    return False
After defining these properties, the spoke can control its accessibility and completeness, but it cannot provide a summary of the values configured within - you must visit the spoke to see how it is configured, which may not be desired. For this reason, an additional property called status exists; this property contains a single line of text with a short summary of configured values, which can then be displayed in the hub under the spoke title.
The status property is defined in the Hello World example add-on as follows:

Example 12. Defining the status Property

@property
def status(self):
    """
    The status property that is a brief string describing the state of the
    spoke. It should describe whether all values are set and if possible
    also the values themselves. The returned value will appear on the hub
    below the spoke's title.

    :rtype: str

    """

    text = self.data.addons.org_fedora_hello_world.text

    # If --reverse was specified in the kickstart, reverse the text
    if self.data.addons.org_fedora_hello_world.reverse:
        text = text[::-1]

    if text:
        return _("Text set: %s") % text
    else:
        return _("Text not set")
After defining all properties described in this chapter, the add-on has full support for the graphical user interface as well as Kickstart. Note that the example demonstrated here is very simple and does not contain any controls; knowledge of Python Gtk programming is required to develop a functional, interactive spoke in the GUI.
One notable restriction is that each spoke must have its own main window - an instance of the SpokeWindow widget. This widget, along with some other widgets specific to Anaconda, is found in the anaconda-widgets package. Other files required for development of add-ons with GUI support (such as Glade definitions) can be found in the anaconda-widgets-devel package.
Once your graphical interface support module contains all necessary methods you can continue with the following section to add support for the text-based user interface, or you can continue with Section 5.7, “Deploying and testing an Anaconda add-on” and test the add-on.
5.6.2.2. Advanced features
The pyanaconda package contains several helper and utility functions and constructs which may be used by hubs and spokes and which have not been covered in the previous section. Most of them are located in pyanaconda.ui.gui.utils.
The sample Hello World add-on demonstrates usage of the englightbox content manager which is also used in Anaconda. This manager can put a window into a lightbox to increase its visibility and focus it and to prevent users interacting with the underlying window. To demonstrate this function, the sample add-on contains a button which opens a new dialog window; the dialog itself is a special HelloWorldDialog inheriting from the GUIObject class, which is defined in pyanaconda.ui.gui.__init__.
The dialog class defines the run method which runs and destroys an internal Gtk dialog accessible through the self.window attribute, which is populated using a mainWidgetName class attribute with the same meaning. Therefore, the code defining the dialog is very simple, as demonstrated in the following example:

Example 13. Defining a englightbox Dialog

# every GUIObject gets ksdata in __init__
dialog = HelloWorldDialog(self.data)

# show dialog above the lightbox
with enlightbox(self.window, dialog.window):
    dialog.run()
The code above creates an instance of the dialog and then uses the enlightbox context manager to run the dialog within a lightbox. The context manager needs a reference to the window of the spoke and to the dialog's window to instantiate the lightbox for them.
Another useful feature provided by Anaconda is the ability to define a spoke which will appear both during the installation and after the first reboot (in the Initial Setup utility described in Section 5.1.2, “Firstboot and Initial Setup”). To make a spoke available in both Anaconda and Initial Setup, you must inherit the special FirstbootSpokeMixIn (or, more precisely, mixin) as the first inherited class defined in the pyanaconda.ui.common module.
If you want to make a certain spoke available only in Initial Setup, you should instead inherit the FirstbootOnlySpokeMixIn class.
There are many more advanced features provided by the pyanaconda package (like the @gtk_action_wait and @gtk_action_nowait decorators), but they are out of scope of this guide. Readers are recommended to go through the installer's sources for examples.

5.6.3. Text User Interface

The third supported interface, after Kickstart and GUI which have been discussed in previous sections, Anaconda also supports a text-based interface. This interface is more limited in its capabilities, but on some systems it may be the only choice for an interactive installation. For more information about differences between the text-based and graphical interface and about limitations of the TUI, see Section 5.1.1, “Introduction to Anaconda”.
To add support for the text interface into your add-on, create a new set of subpackages under the tui directory as described in Section 5.5, “Anaconda Add-on Structure”.
Text mode support in the installer is based on the simpleline utility, which only allows very simple user interaction. It does not support cursor movement (instead acting like a line printer) nor any visual enhancements like using different colors or fonts.
Internally, there are three main classes in the simpleline toolkit: App, UIScreen and Widget. Widgets, which are units containing information to be shown (printed) on the screen, are placed on UIScreens which are switched by a single instance of the App class. On top of the basic elements, there are hubs, spokes and dialogs, all containing various widgets in a way similar to the graphical interface.
For an add-on, the most important classes are NormalTUISpoke and various other classes defined in the pyanaconda.ui.tui.spokes package. All of those classes are based on the TUIObject class, which itself is an equivalent of the GUIObject class discussed in the previous chapter. Each TUI spoke is a Python class inheriting from the NormalTUISpoke class, overriding special arguments and methods defined by the API. Because the text interface is simpler than the GUI, there are only two such arguments:
  • title - determines the title of the spoke, same as the title argument in the GUI
  • category - determines the category of the spoke as a string; the category name is not displayed anywhere, it is only used for grouping

    Note

    Categories are handled differently than in the GUI. [5] It is recommended to assign a pre-existing category to your new spoke. Creating a new category would require patching Anaconda, and brings little benefit.
Each spoke is also expected to override several methods, namely __init__, initialize, refresh, refresh, apply, execute, input, and prompt, and properties (ready, completed, mandatory, and status). All of these have already been described in Section 5.6.2, “Graphical user interface”.
The example below shows the implementation of a simple TUI spoke in the Hello World sample add-on:

Example 14. Defining a Simple TUI Spoke

def __init__(self, app, data, storage, payload, instclass):
    """
    :see: pyanaconda.ui.tui.base.UIScreen
    :see: pyanaconda.ui.tui.base.App
    :param app: reference to application which is a main class for TUI
                screen handling, it is responsible for mainloop control
                and keeping track of the stack where all TUI screens are
                scheduled
    :type app: instance of pyanaconda.ui.tui.base.App
    :param data: data object passed to every spoke to load/store data
                 from/to it
    :type data: pykickstart.base.BaseHandler
    :param storage: object storing storage-related information
                    (disks, partitioning, bootloader, etc.)
    :type storage: blivet.Blivet
    :param payload: object storing packaging-related information
    :type payload: pyanaconda.packaging.Payload
    :param instclass: distribution-specific information
    :type instclass: pyanaconda.installclass.BaseInstallClass

    """

    NormalTUISpoke.__init__(self, app, data, storage, payload, instclass)
    self._entered_text = ""

def initialize(self):
    """
    The initialize method that is called after the instance is created.
    The difference between __init__ and this method is that this may take
    a long time and thus could be called in a separated thread.

    :see: pyanaconda.ui.common.UIObject.initialize

    """

    NormalTUISpoke.initialize(self)

def refresh(self, args=None):
    """
    The refresh method that is called every time the spoke is displayed.
    It should update the UI elements according to the contents of
    self.data.

    :see: pyanaconda.ui.common.UIObject.refresh
    :see: pyanaconda.ui.tui.base.UIScreen.refresh
    :param args: optional argument that may be used when the screen is
                 scheduled (passed to App.switch_screen* methods)
    :type args: anything
    :return: whether this screen requests input or not
    :rtype: bool

    """

    self._entered_text = self.data.addons.org_fedora_hello_world.text
    return True

def apply(self):
    """
    The apply method that is called when the spoke is left. It should
    update the contents of self.data with values set in the spoke.

    """

    self.data.addons.org_fedora_hello_world.text = self._entered_text

def execute(self):
    """
    The excecute method that is called when the spoke is left. It is
    supposed to do all changes to the runtime environment according to
    the values set in the spoke.

    """

    # nothing to do here
    pass

def input(self, args, key):
    """
    The input method that is called by the main loop on user's input.

    :param args: optional argument that may be used when the screen is
                 scheduled (passed to App.switch_screen* methods)
    :type args: anything
    :param key: user's input
    :type key: unicode
    :return: if the input should not be handled here, return it, otherwise
             return True or False if the input was processed succesfully or
             not respectively
    :rtype: bool|unicode

    """

    if key:
        self._entered_text = key

    # no other actions scheduled, apply changes
    self.apply()

    # close the current screen (remove it from the stack)
    self.close()
    return True

def prompt(self, args=None):
    """
    The prompt method that is called by the main loop to get the prompt
    for this screen.

    :param args: optional argument that can be passed to App.switch_screen*
                 methods
    :type args: anything
    :return: text that should be used in the prompt for the input
    :rtype: unicode|None

    """

    return _("Enter a new text or leave empty to use the old one: ")
It is not necessary to override the __init__ method if it only calls the ancestor's __init__, but the comments in the example describe the arguments passed to constructors of spoke classes in an understandable way.
The initialize method sets up a default value for the internal attribute of the spoke, which is then updated by the refresh method and used by the apply method to update Kickstart data. The only differences in these two methods from their equivalents in the GUI is the return type of the refresh method (bool instead of None) and an additional args argument they take. The meaning of the returned value is explained in the comments - it tells the application (the App class instance) whether this spoke requires user input or not. The additional args argument is used for passing extra information to the spoke when scheduled.
The execute method has the same purpose as the equivalent method in the GUI; in this case, the method does nothing.
Methods input and prompt are specific to the text interface; there are no equivalents in Kickstart or GUI. These two methods are responsible for user interaction.
The prompt method should return a prompt which will be displayed after the content of the spoke is printed. After a string is entered in reaction to the prompt, this string is passed to the input method for processing. The input method then processes the entered string and takes action depending on its type and value. The above example asks for any value and then stores it as an internal attribute (key). In more complicated add-ons, you typically need to perform some non-trivial actions, such as parse c as "continue" or r as "refresh", convert numbers into integers, show additional screens or toggle boolean values.
Return value of the input class must be either the INPUT_PROCESSED or INPUT_DISCARDED constant (both of these are defined in the pyanaconda.constants_text module), or the input string itself (in case this input should be processed by a different screen).
In contrast to the graphical mode, the apply method is not called automatically when leaving the spoke; it must be called explicitly from the input method. The same applies to closing (hiding) the spoke's screen, which is done by calling the close method.
To show another screen (for example, if you need additional information which was entered in a different spoke), you can instantiate another TUIObject and call one of the self.app.switch_screen* methods of the App.
Due to restrictions of the text-based interface, TUI spokes tend to have a very similar structure: a list of checkboxes or entries which should be checked or unchecked and populated by the user. The previous paragraphs show a way to implement a TUI spoke where the its methods handle printing and processing of the available and provided data. However, there is a different way to accomplish this using the EditTUISpoke class from the pyanaconda.ui.tui.spokes package. By inheriting this class, you can implement a typical TUI spoke by only specifying fields and attributes which should be set in it. The example below demonstrates this:

Example 15. Using EditTUISpoke to Define a Text Interface Spoke

class _EditData(object):
    """Auxiliary class for storing data from the example EditSpoke"""

    def __init__(self):
        """Trivial constructor just defining the fields that will store data"""

        self.checked = False
        self.shown_input = ""
        self.hidden_input = ""

class HelloWorldEditSpoke(EditTUISpoke):
    """Example class demonstrating usage of EditTUISpoke inheritance"""

    title = _("Hello World Edit")
    category = "localization"

    # simple RE used to specify we only accept a single word as a valid input
    _valid_input = re.compile(r'\w+')

    # special class attribute defining spoke's entries as:
    # Entry(TITLE, ATTRIBUTE, CHECKING_RE or TYPE, SHOW_FUNC or SHOW)
    # where:
    #   TITLE specifies descriptive title of the entry
    #   ATTRIBUTE specifies attribute of self.args that should be set to the
    #             value entered by the user (may contain dots, i.e. may specify
    #             a deep attribute)
    #   CHECKING_RE specifies compiled RE used for deciding about
    #               accepting/rejecting user's input
    #   TYPE may be one of EditTUISpoke.CHECK or EditTUISpoke.PASSWORD used
    #        instead of CHECKING_RE for simple checkboxes or password entries,
    #        respectively
    #   SHOW_FUNC is a function taking self and self.args and returning True or
    #             False indicating whether the entry should be shown or not
    #   SHOW is a boolean value that may be used instead of the SHOW_FUNC
    #
    #   :see: pyanaconda.ui.tui.spokes.EditTUISpoke
    edit_fields = [
        Entry("Simple checkbox", "checked", EditTUISpoke.CHECK, True),
        Entry("Always shown input", "shown_input", _valid_input, True),
        Entry("Conditioned input", "hidden_input", _valid_input,
        lambda self, args: bool(args.shown_input)),
        ]

    def __init__(self, app, data, storage, payload, instclass):
        EditTUISpoke.__init__(self, app, data, storage, payload, instclass)

        # just populate the self.args attribute to have a store for data
        # typically self.data or a subtree of self.data is used as self.args
        self.args = _EditData()

    @property
    def completed(self):
        # completed if user entered something non-empty to the Conditioned input
        return bool(self.args.hidden_input)

    @property
    def status(self):
        return "Hidden input %s" % ("entered" if self.args.hidden_input
                                    else "not entered")

    def apply(self):
        # nothing needed here, values are set in the self.args tree
        pass
The auxiliary class _EditData serves as a data container which is used to store values entered by the user. The HelloWorldEditSpoke class defines a simple spoke with one checkbox and two entries, all of which are instances of the EditTUISpokeEntry class imported as the Entry class). The first one is shown every time the spoke is displayed, the second instance is only shown if the first one contains a non-empty value.
For more information about the EditTUISpoke class, see the comments in the above example.

5.7. Deploying and testing an Anaconda add-on

To test a new add-on, you must load it into the installation environment. Add-ons are collected from the /usr/share/anaconda/addons/ directory in the installation runtime environment; to add your own add-on into that directory, you must create a product.img file with the same directory structure and place it on your boot media.
For specific instructions on unpacking an existing boot image, creating a product.img file and repackaging the image, see Section 2, “Working with ISO Images”.


[1] While Firstboot is a legacy tool, it is still supported because of third-party modules written for it.
[2] In Fedora, the add-on is disabled by default. You can enable it using the inst.kdump_addon=on option in the boot menu.
[3] The gui package may also contain a categories subpackage if the add-on needs to define a new category, but this is not recommended.
[4] an instance of the SpokeWindow widget which is a custom widget created for the Anaconda installer
[5] which is likely to change in the future to sticking to the better (GUI) way