在 spark 数据框中运行 UDF 时，不支持获取 org.apache.spark.sql.Column 类型的架构

Question

我正在尝试连接 spark 数据框中的列数组，我正在通过 spark scala UDF 接收列数组。

这是我的代码：

val aaa = Map(("00","DELHI") -> (List("key1","key2","key3"),List("a")))

   val sampleDf = sparksession.createDataFrame(
      List(("00", "DELHI", "111", "222", "333"), ("00", "SP", "123123123", "231231231", "312312312")
      )).toDF("RecordType", "CITY", "key1", "key2", "key3")  //.printSchema() //.show(100,false)

    val test2 = sampleDf.withColumn("primayKEY",concat(getprimakey(aaa)(col("RecordType"),col("CITY")))).show()//.printSchema()//show(false)


  def getprimakey (mapconfig: Map[(String, String), (List[String], List[String])]) =  udf((rec:String ,layout:String) => {
    println(rec+""+layout)
    val s = mapconfig(rec,layout)._1.map(x => col(x)).toArray//.map(x => col(x))
    s
  })

下面是我遇到的错误

Exception in thread "main" java.lang.UnsupportedOperationException: Schema for type org.apache.spark.sql.Column is not supported
    at org.apache.spark.sql.catalyst.ScalaReflection$.schemaFor(ScalaReflection.scala:733)
    at org.apache.spark.sql.catalyst.ScalaReflection$.schemaFor(ScalaReflection.scala:693)
    at org.apache.spark.sql.catalyst.ScalaReflection$.schemaFor(ScalaReflection.scala:671)
    at org.apache.spark.sql.functions$.udf(functions.scala:3088)
    at com.rupesh.TEST_UDF$.getprimakey(TEST_UDF.scala:29)
    at com.rupesh.TEST_UDF$.main(TEST_UDF.scala:19)
    at com.rupesh.TEST_UDF.main(TEST_UDF.scala)

Answer 1

您只能访问您传递给 UDF 的 UDF 中的字段。因此，您的逻辑需要整行，这可以通过传递 struct("*") 来完成：

def getprimakey(mapconfig: Map[(String, String), (List[String], List[String])]) = udf((rec: String, layout: String, entireRow:Row) => {
  mapconfig.get(rec,layout).map(_._1)
  .map(k => k.map(entireRow.getAs[String](_)))
  .map(_.mkString)
})

sampleDf.withColumn("primayKEY", getprimakey(aaa)(col("RecordType"), col("CITY"), struct("*"))).show() 

+----------+-----+---------+---------+---------+---------+
|RecordType| CITY|     key1|     key2|     key3|primayKEY|
+----------+-----+---------+---------+---------+---------+
|        00|DELHI|      111|      222|      333|111222333|
|        00|   SP|123123123|231231231|312312312|     null|
+----------+-----+---------+---------+---------+---------+