Scala一行代码实现WorldCount

1.辣行代码

lines.flatMap(_.split(" ")).map((_, 1)).groupBy(_._1).map(tuple => (tuple._1, tuple._2.size)).toList.sortBy(_._2)

2.代码分解

package com.w4xj.scala.study0502

object WorldCount {
  def main(args: Array[String]): Unit = {
    val lines = List("scala spark storm hadoop sqoop flume",
      "kafka scala spark oozie scala spark zookeeper",
      "hadoop hive hbase scala")

    //1.扁平化，即切割
    //val result1 = lines.flatMap((str:String ) => str.split(""))
    //简写
    val result1 = lines.flatMap(_.split(" "))

    //2.组成对偶
    //result1.map((str:String) => (str,1))
    //简写
    val result2 = result1.map((_, 1))

    //3.分组,注意元祖的下标是1开始
    //result2.groupBy((tuple: (String, Int)) => tuple._1)
    //简写
    val result3 = result2.groupBy(_._1)

    //4.统计，即reduce阶段的统计
    //val result4 = result3.map((tuple: (String, List[(String, Int)])) => (tuple._1, tuple._2.size))
    //简写,这里的变量不能省略，因为函数体里用了2次
    val result4 = result3.map(tuple => (tuple._1, tuple._2.size))

    //排序，将map转为元祖的List进行排序
    //result4.toList.sortBy((tuple:(String,Int)) => tuple._2)
    //简写
    val result5 = result4.toList.sortBy(_._2)

    //打印：result5 = List((storm,1), (sqoop,1), (oozie,1), (kafka,1), (hive,1), (zookeeper,1), (flume,1), (hbase,1), (hadoop,2), (spark,3), (scala,4))
    println("result5 = " + result5)


    println("=========================================")
    //一条

    val scalaWorldCount = lines.flatMap(_.split(" ")).map((_, 1)).groupBy(_._1).map(tuple => (tuple._1, tuple._2.size)).toList.sortBy(_._2)
    //打印：scalaWorldCount = List((storm,1), (sqoop,1), (oozie,1), (kafka,1), (hive,1), (zookeeper,1), (flume,1), (hbase,1), (hadoop,2), (spark,3), (scala,4))
    println("scalaWorldCount = " + scalaWorldCount)


  }
}