Fill null with previous value not cascading values

[francosang]

Given this code:

val csv = """
        Name,Belongings
        Peter,TV
        ,Phone
        ,Car
        Andy,
        Mary,House
        ,Boat
        ,Dog
        ,Laptop
    """.trimIndent().toInputStream()

DataFrame.readCSV(csv)
    .fillNulls("Name").with { prev()?.get("Name") }
    .forEach { println(it) }

Current output:

{ Name:Peter, Belongings:TV }
{ Name:Peter, Belongings:Phone }
{ Belongings:Car }
{ Name:Andy }
{ Name:Mary, Belongings:House }
{ Name:Mary, Belongings:Boat }
{ Belongings:Dog }
{ Belongings:Laptop }

When the row index 1 (value ,Phone) is reached, the fill with block is executed and the previous row (index 0 Peter,TV) contains the Name as expected.
But it seems that for the next row (index 2 ,Car), the previous row does not contain the value that was filled previously.

Expected output:

{ Name:Peter, Belongings:TV }
{ Name:Peter, Belongings:Phone }
{ Name:Peter, Belongings:Car }
{ Name:Andy }
{ Name:Mary, Belongings:House }
{ Name:Mary, Belongings:Boat }
{ Name:Mary, Belongings:Dog }
{ Name:Mary, Belongings:Laptop }

Version:

implementation("org.jetbrains.kotlinx:dataframe:0.15.0")