dynamically group by columns

[wer vetra]

Similar to this issue, Is it possible to optionally groupby a column of known type (e.g. Option[String])? Is it possible to group by Seq of column name Strings (Seq[String])? And, is it possible to do both (Option[Seq[String]])?

Borrowing from this proposed solution, I can dynamically define the column in the group by, like:



def getSomethingGroupedBy(groupBy: String): Future[Seq[String]] = {
 val query = someTable
 .groupBy(cols => cols.column[String](groupBy)))
 .map{
 case (group, row) => group
 }

 db.run(query.result.transactionally)
}

But if I try to use a groupBy: Seq[String], I get:

Slick does not know how to map the given types.
[error] Possible causes: T in Table[T] does not match your * projection,
[error]  you use an unsupported type in a Query (e.g. scala List),
[error]  or you forgot to import a driver api into scope.
[error]   Required level: slick.lifted.FlatShapeLevel
[error]      Source type: Seq[slick.lifted.Rep[String]]
[error]    Unpacked type: T
[error]      Packed type: G
[error]       .groupBy(cols => groupBy.map(g => cols.column[String](g)))
[error]               ^
[error] one error found


How can I group by an optional and/or arbitrary Seq of string columns that are user defined?