* moved unique value naming to datasource (credit: @bergquist)
* merge rows based on same column-values and empty values
* expanded tests