Make sure data is not duplicated

When descriptorParser parses a descriptor, or consensus document, or any other tor network data file, it generates a digest that is used as id for that data. In many cases also creates some sub records, these do not always include a digest and we should make sure that we are not inserting duplicated data.