It's not exactly clear, but OP is asking about indexing the content of all the documents, not the metadata (e.g. titles etc)