The modern information system cannot live on its own. It looks quite feasible to use Bagri to process CSV or HTML files. Or, to get documents in the system from some external storage (RDBMS, Hadoop, MongoDB or whatever..). The task to integrate Bagri with all possible external systems is out of scope for our team. That is why we’ve decided to provide open APIs for that kind of tasks, so any developer could implement and plug-in integration module to his/her document store and/or format. See our javadocs for parsers and storage APIs to get more details.
See also bagri-extensions project which is already provides some of the storage/format extensions mentioned above. Need a connector to your own store? Fill free to ask for any help in the extension implementation and join the project!