Support a new field in datasets called `join` which will be an array of values of other datasets UUIDs and the result will be a dataset with all the rows and a new column identifying the origin dataset.

So, merging dataset a with values (A, 1) and dataset b with values (B, 1) should result in a new dataset with the following content:

A, 1, a
B, 1, b 

To join datasets the following must be respected:

• Joined datasets must have the exactly same ids and type of columns
• The joiner will not support cache or url/content, it will rely solely on the joined datasets content and configuration

This dataset will be useful for joining Prometheus metrics since PromQL has no support to join queries when using aggregation functions (e.g. vector sum);