测试代码 

package Data_text

import org.apache.flink.streaming.api.TimeCharacteristic
import org.apache.flink.streaming.api.functions.timestamps.BoundedOutOfOrdernessTimestampExtractor
import org.apache.flink.streaming.api.scala._
import org.apache.flink.streaming.api.scala.function.ProcessAllWindowFunction
import org.apache.flink.streaming.api.windowing.assigners.TumblingEventTimeWindows
import org.apache.flink.streaming.api.windowing.time.Time
import org.apache.flink.streaming.api.windowing.windows.TimeWindow
import org.apache.flink.util.Collector

import java.text.SimpleDateFormat
import scala.collection.mutable
import scala.collection.mutable.ListBuffer

object pv {
  def main(args: Array[String]): Unit = {
    //创建流环境
    val env = StreamExecutionEnvironment.getExecutionEnvironment
    //自定义进程为1
    env.setParallelism(1)
    //定义时间语义
    env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime)
    //时间格式转换
    val format = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss")
    //获取数据
    val stream = env.readTextFile("E:\\idea项目\\untitled4\\src\\main\\resources\\UserBehavior.csv")
    //处理数据
    val stream2 = stream
      .map(data=>{
        val arr = data.split(",")
        (arr(0),arr(1),arr(2),arr(3),arr(4).toLong,1)
      })
      .assignAscendingTimestamps(data=>{
              data._5 * 1000L
            })
      .windowAll(TumblingEventTimeWindows.of(Time.hours(1))) //开窗1个小时
      .process(new ProcessAllWindowFunction[(String,String,String,String,Long,Int),String,TimeWindow] {
        override def process(context: Context, elements: Iterable[(String, String, String, String,Long,Int)],
                             out: Collector[String]): Unit = {
          //定义hashMap  
          val hashMap = new mutable.HashMap[String,Int]()
          //遍历数据，将数据装入hashMap  
          for(i <- elements){
            if(hashMap.contains(i._2)){
              // 计算点击量，一条数据一次点击量 
              val sum = hashMap(i._2) + 1  
              hashMap.update(i._2,sum)
            }else{
              hashMap.put(i._2,i._6)
            }
          }
            //定义一个可变的listBuffer，装填数据，自行排序
          val listBuffer = new ListBuffer[(String,Int)]
          for(i <- hashMap){
            listBuffer += i
          }
          val listBuffer2 = listBuffer
            .sortBy(_._2)(Ordering.Int.reverse)
            .take(3)
           //输出 
          out.collect("开始时间"+format.format(context.window.getStart)+"结束时间"+format.format(context.window.getEnd)+"数据："+listBuffer2)

        }
      })


    stream2.print("")
    env.execute("")
  }

}

运行结果