archutil

archutil is an Arch Linux utility that makes it easy to replicate an Arch Linux setup on any other machine in terms of both packages and configuration files. The configuration management functions can actually be used on any Linux distro, but the package management functions can only be used on Arch Linux While there are no plans to port this utility to other platforms, this shouldn't be hard to do with the current design of the tools. You can contact me if you're interested in working on this and pull requests are welcome.

This tool is still in development and is bound to have bugs. Please report any bugs you find in the issues section. There are also many features that could be added to improve the tool that weren't absolutely necessary for the release. If there's a feature you would find useful, please create an issue for it.

Package Management

The first use for archutil is for maintaining a package list so you can keep different Arch installs in sync, quickly replicate an Arch install on another machine, or selectively install specific groups of packages on a machine (i.e. there's no point in installing graphical packages from your local machine to a server running Arch). To list all of the explicitly installed packages on your machine (note that this does not include dependencies), run ./archutil.py list. On your first run, this should essentially match the output of pacman -Qe (minus the base group, but we'll get to that). Place that list of packages into config.py in the packages Python dictionary (i.e. replace package1 and package2 with the packages given to you by archutil). Now if you run ./archutil.py list again, it'll only list the packages you've explicitly installed but haven't added to your list, so in this case you should see nothing. You can try installing a random package at this point and then running the command to see if it works. This helps you easily maintain the package list in config.py. Note that if you put a package group (like gnome or base-devel) in your package list, all of the packages in that group won't be displayed in the output of the list command. The base package group is implicitly placed in your list, as every Arch Linux system should have these packages installed, so you should never see these packages in the output.

This also helps you keep your system lean and minimal. After putting the initial package list in config.py and not running archutil for a while, you may end up with a bunch of new packages on your machine. You might have only been trying these packages out, or maybe you needed a package for something but no longer need it. If you run ./archutil.py list, you can quickly find all the packages you've installed since last running ./archutil.py list, and you can choose to remove those packages or add them to config.py. Thus, you have complete knowledge and control over the packages installed on your system.

You can also organize your packages into categories. One benefit of this is it helps you understand and maintain what packages you have installed. Another (and more important) benefit is that you can install specific categories of packages on certain systems. For example, you wouldn't want to install all the graphical packages from your personal laptop onto a server running Arch. To specify categories in config.py, simply use the category name as the key and a list of packages in that category as an example. You can then refer to these categories with the --categories flag, which we'll use in a bit.

We've talked about listing packages, but how do we actually install them? You can use ./archutil.py install for that. This command first checks if all the packages about to be installed actually exist and tells you if there are any issues. It then installs them with the --needed flag, so only new packages will be installed. If you only want to install certain categories, you can run ./archutil.py install --categories graphical dev networking, which would install all packages in the graphical, dev, and networking categories (note that these are user-defined categories that I made up for this example).

Configuration Management

The second use for archutil is for maintaining your configuration files. I (like many of you) version control my dotfiles and other configuration files. A common issue I run into is forgetting to update my Git repo after making changes to certain files, or making changes on one machine, updating the repo, but pulling the repo on another machine a few days later and not remembering which configuration files to sync. archutil helps to manage both of these issues.

The config_files dictionary in config.py contains a list of configuration files to manage. All configuration files should be stored in a directory called config_files in the same folder as archutil.py (more on customizing this path later). Then the keys of the config_files dictionary are the paths to files in the config_files folders (relative to the config_files folder). The value of each key is the location of the file on the system. The example config.py shows this for a .bashrc file.

To see which files on the system differ from the files in your dotfiles repo, run ./archutil.py config -d. This will print the files that differ. If you also want to see the output of the diff command for each file, run ./archutil.py config -dd.

To install all the configuration files in the system with the files in the repo, run ./archutil.py config -i to install the files. If a file already exists at the path of the system file, a .bak extension will be added to the original file. To update files in the repo with files in the system, run ./archutil.py config -u. This will show the diff for each configuration file and prompt you to update the file in the repo with the file in the system.

archutil will look in a directory named config_files in the same folder as the archutil script by default. If you would like to specify a different folder to search for the config files, you can use the -cd or --configs-dir flags, i.e. ./archutil.py config -cd /path/to/config/files -d. Alternatively, you can define a variable named configs_dir in config.py that contains a path to the configuration file directory. If the path is a relative path, it should be relative to the config.py script, not archutil.py.

Misc features

• By default, archutil looks for config.py in the same directory as the script. If you would like to change this directory, use the -c flag, i.e. ./archutil.py -c /path/to/config.py.
• Note that any of the variables in config.py can be assigned as the output of some function. Thus, if you want to do something like customize the home directory in the paths in the config_files dictionary depending on the user running the script, you can write a Python function to do that
• You can define a required_repos variable in config.py that contains a list of repos that should be enabled in /etc/pacman.conf. Doing this won't enable these repositories, but it will remind you to enable them when installing packages on a new system.
• archutil supports installing packages from the AUR. Define a variable in config.py called pacman and set it to a package manager that can handle packages from the normal repositories and from the AUR, like yaourt. The binary specified in that variable will be used for all operations where pacman would normally have been used, so not all package managers will work. Also, installing packages from the AUR will be slightly slower, as archutil first validates that all packages exist, and currently for AUR packages it needs to run something like yaourt -Ss package_name for each AUR package to make sure it exists.
• archutil attempts to verify all packages that you're about to install actually exist before attempting to install them. This check is usually very quick, except when using a package manager other than pacman, as described above. In these cases, if you would like to skip this verification, you can pass the -s flag to the install subcommand. The package verification for AUR packages will be sped up in the future.