The LIST Aggregate Function for Relational Databases
Witold Litwin · 2003
We propose a new aggregate function. The LIST function aggregates a set of values of one or more attributes into a single value that is internally a list of these values, perhaps ordered. The principle may seem a formal twist, but we show it should be useful in practice. It overcomes important limitations of the current relational systems, due to the use of relations in first normal form, and the separation between the aggregate and the individual data values in the standard SQL. LIST function should be simple to implement, since the relational systems already provide most of the capabilities it requires to the existing aggregate functions.