Hi Denis,
Basically we have hacked Zoo to store it's JSON records in mongoDB... Everything else happens in mysql so if you are using Zoo to persist the document/record then our hack it should work without any refactoring.
We did it this way to provide ourselves with an ability to list, search and filter records in interesting ways. We also did it so that we could use and update the records elsewhere but this is all managed within our own software. It gives us a way to accomplish some of what you did with JBZoo and also to provide us with a way to do other things with the Zoo application data. We did this using a middle tier that maintained it's own indexes to give us an indexing capability.
I realize that if the record is updated outside your app this will mean that we will would have to find a way to trigger the update of the JBZoo indexes.. but that is not why I ask. I do not think that I want to redesign that application...
Right now... I am just thinking how to best use the fantastic front-end capabilities of JBZoo on a new project. For this project I need to link app/zoo data with a workflow/task management system (like redmine) so I need to get the zoo/app data in some common form or otherwise share it in some way. I suppose that an alternative might be some form of REST interface between the two systems so that each can update records in the other but this is slower than updating the datastore directly and can potentially fail.
Kind Regards
Chris.