day5：【刷单行为预警：判断同一个用户点击同一个广告的次数，超过阈值则报警】

*本文用到的数据为ais船舶数据AdClickLog.csv可自取
链接：https://pan.baidu.com/s/1C2CkWzwoM4fhcp3NrRnFJQ?pwd=fx11 
提取码：fx11

案例1：判断同一个用户点击同一个广告的次数，超过阈值则报警

**创建样例类caseClass.scala，创建广告点击样例类

//广告点击样例类
case class AdCase(user:Long,ad:Long,province:String,city:String,time:Long)

import org.apache.flink.api.scala._
import org.apache.flink.api.common.eventtime.WatermarkStrategy
import org.apache.flink.api.common.serialization.SimpleStringSchema
import org.apache.flink.api.common.state.{ValueState, ValueStateDescriptor}
import org.apache.flink.connector.kafka.source.KafkaSource
import org.apache.flink.connector.kafka.source.enumerator.initializer.OffsetsInitializer
import org.apache.flink.streaming.api.functions.KeyedProcessFunction
import org.apache.flink.streaming.api.scala.{DataStream, StreamExecutionEnvironment}
import org.apache.flink.util.Collector

object ClickWarn {
  def main(args: Array[String]): Unit = {
    //【Q1：判断同一个用户点击同一个广告如果超过一百次，我们认为存在刷单行为，发出预警】
    val env: StreamExecutionEnvironment = StreamExecutionEnvironment.getExecutionEnvironment

    val kafkaConsumer = KafkaSource.builder[String]()
      //设置kafka的地址
      .setBootstrapServers("192.168.126.51:9092,192.168.126.52:9092,192.168.126.53:9092")
      //设置组别，用来区分不同的人消费同一个主题的数据
      .setGroupId("paipai")
      //设置主题
      .setTopics("flink")
      //设置是从头消费还是从最新位置开始消费
      .setStartingOffsets(OffsetsInitializer.latest())
      //对value进行 反序列化
      .setValueOnlyDeserializer(new SimpleStringSchema())
      //这个不能丢
      .build()

    val value: DataStream[String] = env
      .fromSource(kafkaConsumer, WatermarkStrategy.noWatermarks(), "kafka")

    value
      .map(x=>{
        //分割数据
      val strings: Array[String] = x.split(",")
        //封装样例类
        AdCase(strings(0).toLong,strings(1).toLong,
          strings(2),strings(3),strings(4).toLong)
      })
      //根据用户和广告进行分组
      .keyBy(x=>(x.user,x.ad))
      .process(new AdWarn(100))
      .print()

    env.execute()

  }
}

//根据传参的方式来传入阈值
class AdWarn(x:Int) extends KeyedProcessFunction[(Long,Long),AdCase,String]{
  //定义1个状态变量
  //状态变量1：同一个用户点击同一个广告的次数
  lazy val count: ValueState[Int] = getRuntimeContext
    .getState(new ValueStateDescriptor[Int]("count", classOf[Int]))

  override def processElement(i: AdCase,
                              context: KeyedProcessFunction[(Long, Long), AdCase, String]#Context,
                              collector: Collector[String]): Unit = {

    //更新次数，来一条数据调用一次processElement函数
    count.update(count.value()+1)
    //x是传入的参数，可以传参
    if (count.value()>x) {
      //在数量>100时输出
      collector
        .collect(s"${i.user}用户点击${i.ad}广告超过100次,我们认为存在刷单行为，发出预警")
      //输出报警信息后就清空状态变量的值
        count.clear()
    }
  }
}