python3-libdnf to python3-libdnf5 migration

[JGoutin]

Hello,

I have the following Python script working with the legacy libdnf.

I want to migrate it to use libdnf5 instead (Current attempt here)

The objective is to list all installed repositories and then list all mirrors for these repositories.

I find that there is a get_mirrors() method on the Repo class here. (Note that this method is not listed in the doc)

But, I don't find the way to get a list of installed Repo instances (Or IDs that I can use with RepoSack.create_repo).

Thanks.

[kontura]

This prints ids and mirrors for all installed and enabled repositories.

import libdnf5

base = libdnf5.base.Base()
base.setup()
repo_sack = base.get_repo_sack()
repo_sack.create_repos_from_system_configuration()
repo_sack.load_repos(libdnf5.repo.Repo.Type_AVAILABLE)

query = libdnf5.repo.RepoQuery(base)
query.filter_enabled(True)

for repo in query:
    print(repo.get_id())
    print(repo.get_mirrors())

You can check out some details about the loading of configured repositories in https://dnf5.readthedocs.io/en/latest/tutorial/bindings/python3/repos.html.

Note that the libdnf5.repo.RepoQuery is not connected to the repoquery command on the CLI, it is a query on all known repositories (Repo objects, not packages contained in those repositories).

(Note that this method is not listed in the doc)

Yea, the docs miss the libdnf5::repo::Repo object completely. Unfortunately a lot of the API is not currently exposed in the docs.

[JGoutin]

That's perfect. Thank you!