Flatten a nested JSON Spark DataFrame using Scala, Spark 2.2.+ — a custom solution.

flatten_df.scala

def flattenDataFrame(df: DataFrame): DataFrame = {
  val fields = df.schema.fields
  val fieldNames = fields.map(x => x.name)

  for (i <- fields.indices) {
    val field = fields(i)
    val fieldType = field.dataType
    val fieldName = field.name
    fieldType match {
      case _: ArrayType =>
        val fieldNamesExcludingArray = fieldNames.filter(_ != fieldName)
        val fieldNamesAndExplode = fieldNamesExcludingArray ++ Array(
          s"explode_outer($fieldName) as $fieldName"
        )
        val explodedDf = df.selectExpr(fieldNamesAndExplode: _*)
        return flattenDataFrame(explodedDf)
      case structType: StructType =>
        val childFieldNames =
          structType.fieldNames.map(childname => fieldName + "." + childname)
        val newFieldNames = fieldNames.filter(_ != fieldName) ++ childFieldNames
        import org.apache.spark.sql.functions.col

        val renamedCols =
          newFieldNames.map { x =>
            col(x.toString).as(x.toString.replace(".", "_"))
          }

        val explodedDf = df.select(renamedCols: _*)
        return flattenDataFrame(explodedDf)
      case _ =>
    }
  }

  df
}

Do you write reverse logic of same. how to convert flatten dataframe to nested dataframe. If you help me with code it will life saver for me

@fahadsiddiqui How to deflatern this df into nested json ..?

is there a way we can modify this code to get the parents as well and not just the final nodes? ex: get columns a, b, a.child, b.child

Hello Fahad, this is interesting
we do have similar one in our website:

https://www.24tutorials.com/spark/flatten-json-spark-dataframe/
Hope this will be helpful.
Thanks.