This package can be used to create BagIt style packages programmatically in Python or from the command line. More about BagIt can be found at:

http://purl.org/net/bagit

bagit.py is a standalone python module that you can drop into your project as needed or you can install globally with:

python setup.py install

From python you can use the bagit module to make a bag like this:

import bagit bag = bagit.make_bag('mydir', {'Contact-Name': 'Ed Summers'})

Or if you've got an existing bag

import bagit bag = bagit.Bag('/path/to/bag')

Or from the command line:

bagit.py --contact-name 'Ed Summers' mydir

If you'd like to generate the checksums using parallel system processes, instead of single process:

bagit.make_bag('mydir', {'Contact-Name': 'Ed Summers'}, processes=4)

or:

bagit.py --processes 4 --contact-name 'Ed Summers' mydir

bag --help will give the full set of options.

Development

% git clone git://github.com/edsu/bagit.git % cd bagit % python test.py

If you'd like to see how increasingl parallelization of bag creation on your system effects the time to create a bag try using the included bench utility:

% ./bench.py

TODO:

• define the return value of make_bag
• add validation
• write manifest to tmp file first before moving stuff around, so that if the manifest generation throws an exception the bag directory will be left untouched

Contact

Ed Summers [email protected]

License

Public Domain http://creativecommons.org/licenses/publicdomain/