时间戳

首页 > 解决方案 > org.apache.flink.table.api.TableException: 结果的 Arity [3] 与请求类型的 number[2] 不匹配

apache-flink - org.apache.flink.table.api.TableException: 结果的 Arity [3] 与请求类型的 number[2] 不匹配

问题描述

我有一个简单的测试用例来学习如何使用 Table API 和 case/when，如下所示：

import org.apache.flink.api.scala.ExecutionEnvironment
import org.apache.flink.table.api.TableEnvironment

import org.apache.flink.api.scala._
case class Person(name: String, age: Int)
object TableTest {
  def main(args: Array[String]): Unit = {
    val env = ExecutionEnvironment.getExecutionEnvironment
    val te = TableEnvironment.getTableEnvironment(env)
    val ds = env.fromCollection(Seq(Person("a",20), Person("b",40), Person("c", 60)))
    te.registerDataSet("person", ds)

    te.toDataSet[Person](table).print()


    val table = te.sqlQuery(
      """
         select name,age,
             case
                when age <= 20 then 'A'
                when age <=40 then 'B'
                when age <= 60 then 'C'
                else 'D'
            end as age_level
            from person
      """.stripMargin(' '))
    te.toDataSet[Person](table).print()
  }
}

运行时遇到如下异常，age_level是计算列，不知道为什么会报错

Exception in thread "main" org.apache.flink.table.api.TableException: Arity [3] of result [ArrayBuffer(String, Integer, String)] does not match the number[2] of requested type [com.flink.table.Person(name: String, age: Integer)].
    at org.apache.flink.table.api.TableEnvironment.generateRowConverterFunction(TableEnvironment.scala:1165)
    at org.apache.flink.table.api.BatchTableEnvironment.getConversionMapper(BatchTableEnvironment.scala:339)
    at org.apache.flink.table.api.BatchTableEnvironment.translate(BatchTableEnvironment.scala:504)
    at org.apache.flink.table.api.BatchTableEnvironment.translate(BatchTableEnvironment.scala:476)
    at org.apache.flink.table.api.scala.BatchTableEnvironment.toDataSet(BatchTableEnvironment.scala:141)
    at com.flink.table.TableTest$.main(TableTest.scala:37)
    at com.flink.table.TableTest.main(TableTest.scala)

标签： apache-flink

解决方案

问题是要转换为 a 的表DataSet[Person]有三个属性：(name, age, age_level)而Personcase 类只有两个字段：(name, age).

你可以实现一个新的案例类

case class PersonWithAgeLevel(name: String, age: Int, age_level: String)

并将表格转换为DataSet[PersonWithAgeLevel].

推荐阅读