Propagate index downstream

natejgardner

If I create a recipe with a single input and output, and the primary index column of the input dataset is still available in the output dataset, I'd like that downstream dataset's index configuration to be set the same as the upstream dataset by default.