`.maxByOption` should return the entire row, not just the column

[mberndt123]

Hi,

In Scala collections, the maxByOption returns the entire object:

scala> Seq((1, 2), (3, 4)).maxByOption(_._1)
val res0: Option[(Int, Int)] = Some((3,4)) // note: it returns (3, 4), not just 3

In ScalaSQL it doesn't. When you call maxByOption, the returned value is only the column. The reason is likely that there could be more than one row where the column reaches its maximum, so it's not obvious which one to choose.
I can see two ways to deal with this:

1. choose an arbitrary one, similar to what .sortBy(_.bla).take(1).singleOption would do
2. raise an error when there's more than one candidate. Note that this error can be avoided by setting a unique constraint on the column, but that is outside the scope of ScalaSQL.

[ETCHKILI]

The semantics of these two functions are different. In Scalasql, maxBy is mapped to sql aggregate function MAX. Thus it returns only one column. There is no simple alternative to Scala maxBy in sql (I'm not pretty sure about this), so i think for a dsl to manipulate sql, this is acceptable.

[mberndt123]

In it's current form, .maxByOption(f) is equivalent to .map(f).maxOption. So… maybe we just don't need it? If it can't be made to work the way it does in Scala collections, then I'm not convinced this particular shorthand is worth it…

[ETCHKILI]

Yeah... Maybe the name is a bit misleading. But considering that many programs may already rely on this function, and because there is type checking here, it will not cause too much trouble overall, and it is unlikely to be deleted.