3.2、spark集群运行应用之第三方jar的处理方式

spark-submit --class a --jars fastjson-1.2.47.jar --master spark://s101:7077 myspark.jar temptags.txt

spark-shell --master spark://s101:7077 --jars fastjson-1.2.47.jar　　//该jar在本地

import java.util._
import scala.collection.JavaConversions._
import scala.collection.JavaConverters._
import com.alibaba.fastjson._
def pp(line: String)={　　//解析方法
    val list = new ArrayList[String]

    val jsonObject = JSON.parseObject(line)
    val extInfoList = jsonObject.getJSONArray("extInfoList")

    if (extInfoList != null && extInfoList.size != 0) {
        for (o <- extInfoList) {
            val jo = o.asInstanceOf[JSONObject]
            if (jo.get("title") == "contentTags") {
                val values = jo.getJSONArray("values")
                for (value <- values) {
                    list.add(value.toString)
                }
            }
        }
    }
    list
}
val rdd1 = sc.textFile("myspark/temptags.txt")
val rdd2 = rdd1.map(s => {val sp = s.split("\t");val lst = pp(sp(1));(sp(0), lst)}).filter(_._2.size() > 0)
val rdd3 = rdd2.flatMapValues(_.asScala).map(t=>((t._1,t._2),1)).reduceByKey((a,b)=>a+b).groupBy(_._1._1).mapValues(_.map(t=>(t._1._2,t._2)))
val rdd4 = rdd3.mapValues(_.toList.sortBy(-_._2)).sortBy(-_._2(0)._2)
val rdd5 = rdd4.collect()
rdd5.foreach(println)