Split by operation on "normal classes" could be clearer

[Jolanrensen]

I was working with a LocalDateTime which I wanted to split up into a LocalDate and a LocalTime column. I found the split operation, but it wasn't immediately obvious how to work with this. For splitting a column up based on String delimiters, something like

df.split { stringCol }.by { it.split(",") }.into("first", "second")

is clear and works well. However, when trying to split up a LocalDateTime:

df.split { localDateTimeCol }.by { listOf(it.date, it.time) }.into("date", "time")

does not seem obvious.

Something like

df.split { localDateTimeCol }.by { it.date and it.time }.into("date", "time")

would improve it a bit, and

df.split { localDateTimeCol }.by { 
  it.date into "date"
  it.time into "time"
}

even more.

(Btw, since Split is a public data class, you're able to call df.split { myCol }.copy { } which looks like it should be prohibited)

[Jolanrensen]

May be related to #1749