package test
import java.util.Properties
import org.apache.spark.SparkConf
import org.apache.spark.SparkContext
import org.apache.spark.sql.{SQLContext, SaveMode}
import org.apache.spark.streaming.Seconds
import org.apache.spark.streaming.StreamingContext
object demo9 {
def main (args : Array[String]) {
val sparkConf = new SparkConf().setMaster("local[2]").setAppName("logapp")
val ssc = new StreamingContext(sparkConf, Seconds(10))
val lines = ssc.socketTextStream("localhost", 9999).map(x =>x.split(" "))
lines.foreachRDD(rdd => {
val sqlContext = SQLContextSingleton.getInstance(rdd.sparkContext)
import sqlContext.implicits._
//构造case class: DapLog,提取日志中相应的字段
val logDataFrame = rdd.map(w => DapLog(w(0).toInt,w(1),w(2),w(3)+" "+w(4))).toDF()
//注册为tempTable
//logDataFrame.registerTempTable("log")
logDataFrame.createOrReplaceTempView("log")
//查询该批次的字段 to_timestamp($"event_time", "MM/dd/yyyy HH:mm:ss")
val logCountsDataFrame = sqlContext.sql("SELECT login_id,user,event_name,to_timestamp(event_time, 'yyyy-MM-dd HH:mm:ss') as event_time,lead(event_time,1,null) over(partition by login_id order by login_id ASC) as nextline_time FROM log")
//打印查询结果
val countN = logCountsDataFrame.count()
if (countN != 0){
logCountsDataFrame.show()
//下面保存会出错,我注释掉了/Users/huiliyang/streaming
//logCountsDataFrame.write.json("/Users/huiliyang/streaming/cc")
//logCountsDataFrame.write.parquet("/Users/huiliyang/streaming/bb")
val prop = new Properties()
prop.put("user", "root")
prop.put("password", "yh200888")
prop.put("driver","com.mysql.jdbc.Driver")
logCountsDataFrame.write.mode(SaveMode.Append).jdbc("jdbc:mysql://localhost:3306/school?useUnicode=true&characterEncoding=utf8", "log", prop)
}
})
ssc.start()
ssc.awaitTermination()
}
}
case class DapLog(login_id:Int, user:String, event_name:String, event_time:String)
object SQLContextSingleton {
@transient private var instance: SQLContext = _
def getInstance(sparkContext: SparkContext): SQLContext = {
if (instance == null) {
instance = new SQLContext(sparkContext)
}
instance
}
}