This python script retrieves an atom feed from Evergreen ILS (http://evergreen-ils.org/), and extracts bibliographic and holdings data into the Exhibit JSON format (http://simile-widgets.org/exhibit/).
I have tested this with Exhibit 3. Here's one! http://www.sandbox.lbcc.linnlibraries.org/newbooks/sandbergja/evergreen2exhibit
Get the JSON data needed to design beautiful visualizations of new books in an Evergreen ILS catalog
PythonMIT