Usage
Repod offers the command-line interface (CLI) tool repod-file (for further
information refer to its manual: repod-file) to interact with package
files and repositories.
Note
The repod-file tool is currently still limited in scope. It can
import packages to existing repositories
update packages in existing repositories
write repository sync databases
inspect package files
write JSON schema used by repod
create archives for package files and signatures
The tool can be used per-user (reading configuration from
$XDG_CONFIG_DIR/repod/repod.conf and
$XDG_CONFIG_DIR/repod/repod.conf.d/ and storing data below
$XDG_STATE_DIR/repod/) or system-wide (reading configuration from
/etc/repod.conf and /etc/repod.conf.d/ and storing data below
/var/lib/repod/).
By default repod-file can be used without a configuration file, as it will
assume a set of defaults (e.g. repository location, default repository name,
default repository architecture). However, it is advisable to create a
configuration file (see repod.conf), as it offers full control over
everything that repod has to offer.
Note
In the future repod will be extended by a web service, which exposes an API (see milestone 0.8.0). This will likely change how interaction with the data maintained by repod takes place.
Using repod-file without a configuration file automatically creates a
repository with the name default and the CPU architecture any.
Importing packages
Packages can be imported to the default repository using
repod-file repo importpkg package_a-1.0.0-1-any.pkg.tar.zst default
Assuming that the pkgbase (see package splitting) of
package_a-1.0.0-1-any.pkg.tar.zst is package_a, the above command will
import the package to the repository named default, write the Sync Database files for it and create the following directory structure:
/home/user/.local/state/repod
├── archive
│ ├── package
│ │ └── p
│ │ └── package_a
│ │ └── package_a-1.0.0-1-any.pkg.tar.zst
│ └── source
├── data
│ ├── pool
│ │ ├── package
│ │ │ └── default
│ │ │ └── package_a-1.0.0-1-any.pkg.tar.zst
│ │ └── source
│ │ └── default
│ └── repo
│ ├── package
│ │ └── default
│ │ └── any
│ │ ├── default.db -> default.db.tar.gz
│ │ ├── default.db.tar.gz
│ │ ├── default.files -> default.files.tar.gz
│ │ ├── default.files.tar.gz
│ │ └── package_a-1.0.0-1-any.pkg.tar.zst -> ../../../../pool/package/default/package_a-1.0.0-1-any.pkg.tar.zst
│ └── source
│ └── default
│ └── any
└── management
└── default
└── any
└── default
└── package_a.json
Note
The data directory (e.g. $XDG_STATE_DIR/repod/data/) is intended to be
exposed by a webserver, as it represents the Binary Repository. The
management directory (e.g. $XDG_STATE_DIR/repod/management/ on the other
hand is not supposed to be exposed publicly, as it represents the
Management Repository and will gain the ability to expose itself using
a git backend in the future (see milestone 0.4.0).
Writing sync databases
The writing of Sync Database files (as is done when Importing packages), can also be triggered manually.
repod-file repo writedb default
The above command creates the following directory structure (assuming no packages are present):
/home/user/.local/state/repod
├── archive
│ ├── package
│ └── source
├── data
│ ├── pool
│ │ ├── package
│ │ │ └── default
│ │ └── source
│ │ └── default
│ └── repo
│ ├── package
│ │ └── default
│ │ └── any
│ │ ├── default.db -> default.db.tar.gz
│ │ ├── default.db.tar.gz
│ │ ├── default.files -> default.files.tar.gz
│ │ └── default.files.tar.gz
│ └── source
│ └── default
│ └── any
└── management
└── default
└── any
└── default
Using repositories
Users of the pacman package manager can add a repository maintained by repod to their pacman.conf (see repository sections for details).
We will assume, that repod’s data directory is exposed at
https://domain.tld. The following addition to /etc/pacman.conf would then
allow users to install package_a from the default repository:
[default]
Server = https://domain.tld/repo/package/$repo/$arch