Gopi,
A perhaps better question is: what are you really trying to do?
I ask because it appears you are thinking of Couchbase views like they are SQL indices (with which they share some behavior). Thinking this way both underutilizes the view's capabilities and will keep you from learning how to model your data as appropriate for a document database. A document database with nested documents is not a relational database; one models one's data differently.
Why do you want a dynamic view? Dynamism always has a cost that exceeds the cost of static, compiled/indexed, routine behavior. Why do you want to pay that cost?
The short answer is that it isn't practically possible to create a dynamic view. The long answer is you can create such a view yourself by reading every document. You typically don't want to do this. In my case, my test database has almost 100M documents and exceeds 90GB in size. Just reading every bit of it at wireline speeds takes over 30 minutes (datapoint taken from the time it takes CB to warm up on this DB from a cold start). IOW, I could never afford to pay the dynamic cost. If your database is of any size, I doubt you can either.