Dataframe 到 RDD 的一段代码不起作用

Question

I am trying to read each row of dataframe and convert the row data into custom bean class. But the problem here is, the code is not getting executed. To check, I have written multiple print statement but none of the print statement present inside df.rdd.map{row=>} executed as if the complete block of code is escaped.

code snippet:

 print("data frame:", df.show()). 

 df.rdd.map(row => {
   // Debugging
   println("Debugging")

  if(row.isNullAt(0)) {
    println("null data")
  } else {
    println(row.get(0).toString)
  }

  val employeeJobData = new EmployeeJobData

  if(row.get(0).toString == null || row.get(0).toString.isEmpty){
    employeeJobData.setEmployeeId("NULL_KEY_VALUE")
  } else {
    employeeJobData.setEmployeeId(row.get(0).toString)
  }
  employeeJobDataList.add(employeeJobData)
  } )

output of df.show() :

   |employee_id|employee_name|employee_email|paygroup|level|dept_id|
   +-----------+-------------+--------------+--------+-----+-------+
   |13         |         null|          null|    null| null|   null|
   |14         |         null|          null|    null| null|   null|
   |15         |         null|          null|    null| null|   null|
   |16         |         null|          null|    null| null|   null|
   |17         |         null|          null|    null| null|   null|
   +-----------+-------------+--------------+--------+-----+-------+

Answer 1

You can remove unnecessary code as below and get java.util.List[EmployeeJobData] as below

import java.util

object MapToCaseClass {

  def main(args: Array[String]): Unit = {
    val spark = Constant.getSparkSess;

    import spark.implicits._

    val df  = List((12,"name","email@email.com","paygroup","level","dept_id")).toDF()
    val employeeList : util.List[EmployeeJobData] = df
      .map(row => {
        val id = if (null == row.getString(0) || "null".equals(row.getString(0)) || row.getString(0).trim.isEmpty) {
          "NULL_KEY_VALUE"
        } else {
          row.getString(0)
        }
        EmployeeJobData(id, row.getString(1), row.getString(2),
          row.getString(3), row.getString(4), row.getString(5))
      })
      .collectAsList
  }

}

case class EmployeeJobData(employee_id: String, employee_name: String,employee_email: String,paygroup: String,
                           level: String,dept_id: String)

The above can be improved more by just setting the data type of employee_id and dept_id (ie if its numeric) to Long . This "null".equals and .isEmpty() can be avoided for employee_id and code can be further reduced.