spark提示Caused by: java.lang.ClassCastException: scala.collection.mutable.WrappedArray$ofRef cannot be cast to [Lscala.collection.immutable.Map;

spark提示Caused by: java.lang.ClassCastException: scala.collection.mutable.WrappedArray$ofRef cannot be cast to [Lscala.collection.immutable.Map;

起因

编写了一个处理两列是否相等的UDF，这两列的数据结构是一样的，但是结构比较复杂，如下：

|-- list: array (nullable = true)
 |    |-- element: map (containsNull = true)
 |    |    |-- key: string
 |    |    |-- value: array (valueContainsNull = true)
 |    |    |    |-- element: struct (containsNull = true)
 |    |    |    |    |-- Date: integer (nullable = true)
 |    |    |    |    |-- Name: string (nullable = true)
 |-- list2: array (nullable = true)
 |    |-- element: map (containsNull = true)
 |    |    |-- key: string
 |    |    |-- value: array (valueContainsNull = true)
 |    |    |    |-- element: struct (containsNull = true)
 |    |    |    |    |-- Date: integer (nullable = true)
 |    |    |    |    |-- Name: string (nullable = true)

也就是说Array里嵌着Map，Map里还嵌着一个Array，只能依次去比较，编写的UDF如下：

case class AppList(Date: Int, versionCode: Int, Name:String)

    def isMapEqual(map1: Map[String, Array[AppList]], map2:Map[String, Array[AppList]]): Boolean = {
      try{
        if (map1.size != map2.size){
          return false
        } else{
          for ( x <- map1.keys){
            if (map1(x) != map2(x)){
              return false
            }
          }
          return true
        }
      } catch {
        case e: Exception => false
      }
    }

    def isListEqual(list1: Array[Map[String, Array[AppList]]], list2:Seq[Map[String, Seq[AppList]]]): Boolean = {
      try {
        if (list1.length != list2.length){
           return false
        } else if (list1.length == 0 || list2.length == 0){
          return false
        } else {
          return isMapEqual(list1(0), list2(0))
        }
      } catch {
        case e: Exception => false
      }
    }

    val isColumnEqual = udf((list1: Array[Map[String, Array[AppList]]], list2:Array[Map[String, Array[AppList]]]) => {
      isListEqual(list1, list2)
    })

然后我就贴到spark-shell里去执行了下面语句：

val dat = df.withColumn("equal", isColumnEqual($"list1", $"list2"))
dat.show()

此时就出现了如下错误：

Caused by: org.apache.spark.SparkException: Failed to execute user defined function($anonfun$1: (array<map<string,array<struct<Date:int,Name:string>>>>, array<map<string,array<struct<Date:int,Name:string>>>>) => boolean)
  at org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIterator.processNext(Unknown Source)
  at org.apache.spark.sql.execution.BufferedRowIterator.hasNext(BufferedRowIterator.java:43)
  at org.apache.spark.sql.execution.WholeStageCodegenExec$$anonfun$8$$anon$1.hasNext(WholeStageCodegenExec.scala:377)
  at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:231)
  at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:225)
  at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
  at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
  at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
  at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:323)
  at org.apache.spark.rdd.RDD.iterator(RDD.scala:287)
  at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87)
  at org.apache.spark.scheduler.Task.run(Task.scala:99)
  at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:322)
  at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
  at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
  at java.lang.Thread.run(Thread.java:745)
Caused by: java.lang.ClassCastException: scala.collection.mutable.WrappedArray$ofRef cannot be cast to [Lscala.collection.immutable.Map;
  at $anonfun$1.apply(<console>:42)
  ... 16 more

解决办法

所谓的解决办法，自然是去谷歌了…

在这里看到，说把Array改成Seq就好了，囧，尝试了一下，果然就好了

原因

这里说：

So it looks like the ArrayType on Dataframe "idDF" is really a WrappedArray and not an Array - So the function call to "filterMapKeysWithSet" failed as it expected an Array but got a WrappedArray/ Seq instead (which doesn‘t implicitly convert to Array in Scala 2.8 and above).

意思是，此Array非Scala中的原生Array，而是封装了一下的Array（有错的一定指出来，我都没写过Scala，慌

参考

• https://stackoverflow.com/questions/40199507/scala-collection-mutable-wrappedarrayofref-cannot-be-cast-to-integer
• https://stackoverflow.com/questions/40764957/spark-java-lang-classcastexception-scala-collection-mutable-wrappedarrayofref