이 콘텐츠는 선택한 언어로 제공되지 않습니다.
5. Developing Installer Add-ons
5.1. Introduction to Anaconda and Add-ons 링크 복사링크가 클립보드에 복사되었습니다!
5.1.1. Introduction to Anaconda 링크 복사링크가 클립보드에 복사되었습니다!
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
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.
5.1.2. Firstboot and Initial Setup 링크 복사링크가 클립보드에 복사되었습니다!
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 링크 복사링크가 클립보드에 복사되었습니다!
%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 링크 복사링크가 클립보드에 복사되었습니다!
- The Anaconda page on Fedora Project Wiki contains provides more information about the installer.
- Information about development of Anaconda into its current version is available at the Anaconda/NewInstaller Wiki page.
- The Kickstart Installations chapter of the Red Hat Enterprise Linux 7 Installation Guide provides full documentation of Kickstart, including a list of all supported commands and options.
- The Installing Using Anaconda chapter of the Red Hat Enterprise Linux 7 Installation Guide describes the installation process in the graphical and text user interfaces.
- For information about tools used for after-installation configuration, see Initial Setup and Firstboot.
5.2. Architecture of Anaconda 링크 복사링크가 클립보드에 복사되었습니다!
pykickstart
- used to parse and validate Kickstart files and also to provide a data structure which stores values which drive the installationyum
- the package manager which handles installation of packages and resolving dependenciesblivet
- originally split from the anaconda package as pyanaconda.storage; used to handle all activities related to storage managementpyanaconda
- 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 functionspython-meh
- contains an exception handler which gathers and stores additional system information in case of a crash and passes this information to thelibreport
library, which itself is a part of the ABRT Project.
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.
/root/
directory on the installed system; therefore the installation can be replicated automatically by reusing this automatically generated Kickstart file.
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
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 링크 복사링크가 클립보드에 복사되었습니다!
- 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
Figure 2. Diagram of the hub and spoke model
Note
- 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).
- 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)
5.4. Threads and Communication 링크 복사링크가 클립보드에 복사되었습니다!
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.
@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.
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.
progressQ
which serves as a medium to transfer installation progress updates.
5.5. Anaconda Add-on Structure 링크 복사링크가 클립보드에 복사되었습니다!
__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.
_
) 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
__init__.py
file in each directory. Directories missing this file are not considered valid Python packages.
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]
ks/
, gui/
and tui/
directories can contain Python modules with any name.
Example 2. Sample Add-on Structure
5.6. Writing an Anaconda add-on 링크 복사링크가 클립보드에 복사되었습니다!
5.6.1. Kickstart Support 링크 복사링크가 클립보드에 복사되었습니다!
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
.
%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.
Example 3. Using an Add-on in a Kickstart File
AddonData
. This class is defined in pyanaconda.addons and represents an object for parsing and storing data from a Kickstart file.
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.
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
Example 4. Using handle_header and handle_line
__all__
variable which is necessary to prevent Anaconda's collect method from taking the AddonData
class instead of add-on specific HelloWorldData
.
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
.
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
.
setup
- called before the installation transaction starts and used to make changes to the installation runtime environmentexecute
- called at the end of the transaction and used to make changes to the target system
Example 5. Importing the setup and execute Methods
setup
and execute
methods included is below:
Example 6. Using the setup and execute Methods
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.
__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
.
__str__
method will be similar to the following example:
Example 7. Defining a __str__ Method
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 링크 복사링크가 클립보드에 복사되었습니다!
Note
SpokeWindow
.
5.6.2.1. Basic features 링크 복사링크가 클립보드에 복사되었습니다!
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”.
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)uiFile
- contains the name of the.glade
filecategory
- contains the class of the category the spoke belongs toicon
- contains the identifier of the icon that will be used for the spoke on the hubtitle
defines the title that will be used for the spoke on the hub
Example 8. Defining Attributes Required for the Normalspoke Class
__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.
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.
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.
__init__
method and the initialize
method.
__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.
__init__
method):
Example 9. Defining the __init__ and initialize Methods
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.
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
.
__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.
__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 theself.data
structure are displayedapply
- called when the spoke is left and used to store values from UI elements back into theself.data
structureexecute
- called when the spoke is left and used to perform any runtime changes based on the new state of the spoke
Example 10. Defining the refresh, apply and execute Methods
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 completedmandatory
- 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)
text
attribute of the HelloWorldData
class:
Example 11. Defining the ready, completed and mandatory Methods
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.
status
property is defined in the Hello World example add-on as follows:
Example 12. Defining the status Property
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.
5.6.2.2. Advanced features 링크 복사링크가 클립보드에 복사되었습니다!
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
.
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__
.
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
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.
FirstbootSpokeMixIn
(or, more precisely, mixin) as the first inherited class defined in the pyanaconda.ui.common
module.
FirstbootOnlySpokeMixIn
class.
@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 링크 복사링크가 클립보드에 복사되었습니다!
tui
directory as described in Section 5.5, “Anaconda Add-on Structure”.
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.
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.
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 thetitle
argument in the GUIcategory
- determines the category of the spoke as a string; the category name is not displayed anywhere, it is only used for groupingNote
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.
__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”.
Example 14. Defining a Simple TUI Spoke
__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.
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.
execute
method has the same purpose as the equivalent method in the GUI; in this case, the method does nothing.
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.
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.
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).
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.
TUIObject
and call one of the self.app.switch_screen*
methods of the App
.
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
_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.
EditTUISpoke
class, see the comments in the above example.
5.7. Deploying and testing an Anaconda add-on 링크 복사링크가 클립보드에 복사되었습니다!
/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.
product.img
file and repackaging the image, see Section 2, “Working with ISO Images”.
inst.kdump_addon=on
option in the boot menu.