12、Spark_RDD算子——MapPartitionWithIndex

一、SparkUtils工具类

import org.apache.spark.{SparkConf, SparkContext}

object SparkUtils {
  /**
   *  默认的master url路径
   */
  val DEFAULT_MASTER = "local[*]"
  /**
   * 默认master为local[*]的获取sparkContext
   */
  def getSparkContext(appName:String):SparkContext = getSparkContext(appName, DEFAULT_MASTER)
  def getSparkContext(appName:String, master:String):SparkContext = new SparkContext(new SparkConf().setAppName(appName).setMaster(master))
  /**
   * 释放sparkContext
   */
  def close(sc:SparkContext) = if(sc != null) sc.stop()
}

二、日志工具

import org.apache.log4j.{Level, Logger}

trait LoggerTrait {
  Logger.getLogger("org.apache.spark").setLevel(Level.WARN)
  Logger.getLogger("org.apache.hadoop").setLevel(Level.WARN)
  Logger.getLogger("org.spark_project").setLevel(Level.WARN)

}

三、Spark算子MapPartitionWithIndex

import cn.qphone.spark.common.LoggerTrait.LoggerTrait
import cn.qphone.spark.common.Utils.SparkUtils
import org.apache.spark.rdd.RDD

object Deom12_MapPartitionWithIndex extends LoggerTrait{
  def main(args: Array[String]): Unit = {
    //1.sparkcontext获取
    val sc = SparkUtils.getSparkContext("Deom12_MapPartitionWithIndex")
    //2.数据
    val list = 1.to(10)
    //3.加载RDD
    val listRDD: RDD[Int] = sc.parallelize(list,1)
    //4.作用MapPartitions分批处理，并不是单个处理，性能要高一些
    val mapRDD: RDD[Int] = listRDD.mapPartitionsWithIndex{
      case (partition,iterator) =>{
        println(s"${partition}.data is ${iterator.mkString("[",",","]")}")
        val ints: Iterator[Int] = iterator.map(_ * 2)
        ints
      }
    }
    //5.打印
    mapRDD.foreach(println)
    //6.释放资源
    SparkUtils.close(sc)
  }
}