This script offers a convenient interface to the tarsnap client. It does not replace it, it just makes certain things more convenient.

1. Create a directory somewhere that contains one directory entry for each archive. Each directory entry may be either a symbolic link or a directory. If it is a directory, it may contain files, symbolic links and subdirectories. You can exclude files within a directory from an archive via the .backup.py.rc config file.
2. From time to time (perhaps via a cronjob), you run backup.py store, which will create new tarsnap archives by appending the current date (yyyy-mm-dd) to each archive name.

To get started, just run backup.py store. You will be told off for not having a .backup.py.rc, and one will be created for you. Customize it by setting your directory (the one containing an entry for each archive) and optional exclusions (corresponding to tarsnap --exclude), and you're ready to go.

Use backup.py view to see at which dates each archive was archived.

If an archive is stored twice in a day, you end up with e.g. foo_2012-03-20 and foo_2012-03-20.1.

Use backup.py list as a convenient way of saying tarsnap --list-archives | sort | grep.

Finally, there's backup.py rename which renames an archive (this operates on the remote archives with their full names, including the date suffix).