Spark Transformation算子-＞mapPartitions

与 map 类似，遍历的单位是每个 partition 上的数据。相对于map算子时一个高性能的算子。

1. java

package transformations;

import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaRDD;
import org.apache.spark.api.java.JavaSparkContext;

import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;


public class MapPartitionsTest {
    public static void main(String[] args) {
        JavaSparkContext context = new JavaSparkContext(
                new SparkConf()
                        .setMaster("local")
                        .setAppName("mapPartitions")
        );
        context.setLogLevel("Error");
        JavaRDD rdd = context.parallelize(Arrays.asList("a", "b", "c", "e", "f", "g"),2);
        List list = new ArrayList<>();
        JavaRDD rdd2 = rdd.mapPartitions(e -> {
            while (e.hasNext())
                list.add(e.next() + "#");
            return list.iterator();
        });
        rdd2.foreach(e-> System.out.print(e+"t"));
    }
}

2. scala

package transformation

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

import scala.collection.mutable.ListBuffer


object MapPartitionTest {
  def main(args: Array[String]): Unit = {
    val context = new SparkContext(
      new SparkConf()
        .setMaster("local")
        .setAppName("mapPartition")
    )
    context.setLogLevel("Error")
    val rdd = context.parallelize(List[String]("a", "b", "c", "e", "f", "g"),2)
    val array: Array[String] = rdd.mapPartitions(e => {
      val buff = new ListBuffer[String]()
      while (e.hasNext) {
        buff.append(e.next() + "# ")
      }
      buff.iterator
    }).collect()
    array.foreach(print)
  }
}