A file pruning utility

Prunehild takes a list of files, and based on a given schedule, it will prune this list based on their date. Different actions can be executed for the pruned files, the most common action would probably be "delete".

The schedule will decide which ones to keep on which ones to delete.

The idea is to use this for example in a directory that is full of backup files. The older they are, the more of them you can 'forget'. Like

backup-sqldump-1.gz
backup-sqldump-2.gz
backup-sqldump-3.gz
backup-sqldump-4.gz
[...]

and remove some of them so the size of the directory stays approximately constant. In this example, you could keep putting sql backup files to the directory and use prunehild to prune them.

The names of the files do not matter, only the date is relevant.

In a schedule you can define how many files you want to keep per interval. A schedule definition of:

"myschedule" : [ {"Day":1}, {"Week":2}, {"Month":1}, {"Long":1} ],

would ensure that as many files that fit into the next bigger interval will be kept. The Long interval is so you can keep some files forever.

Let's demonstrate this with an example:

...

Possible actions you can run on the pruned files:

Print the names of the obsoleted files to stdout. Use this as a dry run option or pipe the output to another command.

Removes the files from disk. Use with care.

Print the names of the files that would be unaffected by pruning. This can be seen as a dry run to see which files will stay after prunhild delete.

Move the files into another directory.

Compress the files, retaining the modification date.