Is there a way to get all the tokens in the term vector of an index

[tyka]

Because I have a large number of documents, retrieving for each document seem to be very inefficient. But it looks like that's the best option available.