sparkstreaming的reduceByKeyAndWindow窗口函数的实现搜索热词排名

 代码实现如下：

import org.apache.spark.SparkConf
import org.apache.spark.streaming.dstream.{DStream, ReceiverInputDStream}
import org.apache.spark.streaming.{Seconds, StreamingContext}

//搜索热词前三，实现实时计算
object WindowScala {
  def main(args: Array[String]): Unit = {


    val conf = new SparkConf().setMaster("local[2]").setAppName("WindowScala")
    val ssc: StreamingContext = new StreamingContext(conf, Seconds(1))
    //获取数据，数据格式为（客户识别信息 搜索内容）
    val ReceiverDStreamInput: ReceiverInputDStream[String] = ssc.socketTextStream("Hadoop01", 9999)
    //将数据进行转化，转化成（搜索内容，1）
    val Dstream = ReceiverDStreamInput.map(line => (line.split(" ")(1), 1))
    //    对数据做reduceByKeyAndWindow操作，对相同的key，按照value求和，窗口长度为30s，每隔10s生成一个窗口
    val windowResults: DStream[(String, Int)] = Dstream.reduceByKeyAndWind