spark之 Idea操作

一、pom.xml

<!-- spark -->
    <dependency>
      <groupId>org.apache.spark</groupId>
      <artifactId>spark-core_2.11</artifactId>
      <version>2.3.4</version>
    </dependency>
    <dependency>
      <groupId>org.apache.spark</groupId>
      <artifactId>spark-sql_2.11</artifactId>
      <version>2.3.4</version>
    </dependency>
    <!-- log4j 1.2.17 -->
    <dependency>
      <groupId>log4j</groupId>
      <artifactId>log4j</artifactId>
      <version>1.2.17</version>
    </dependency>
    <!-- https://mvnrepository.com/artifact/com.google.guava/guava -->
    <dependency>
      <groupId>com.google.guava</groupId>
      <artifactId>guava</artifactId>
      <version>14.0.1</version>
    </dependency>

因为需要打包scala的代码和maven胖包：

<build>
          <!-- 缩短jar包名字 -->
          <finalName>myspark</finalName>
          <plugins>
            <!-- scala -->
            <plugin>
              <groupId>org.scala-tools</groupId>
              <artifactId>maven-scala-plugin</artifactId>
              <version>2.15.2</version>
              <executions>
                <execution>
                  <goals>
                    <goal>compile</goal>
                    <goal>testCompile</goal>
                  </goals>
                </execution>
              </executions>
            </plugin>
            <!-- 胖包 -->
            <plugin>
              <groupId>org.apache.maven.plugins</groupId>
              <artifactId>maven-assembly-plugin</artifactId>
              <version>2.5.5</version>
              <configuration>
                <archive>
                  <manifest>
                    <mainClass>com.njbdqn.MySpark</mainClass>
                  </manifest>
                </archive>
                <descriptorRefs>
                  <descriptorRef>jar-with-dependencies</descriptorRef>
                </descriptorRefs>
              </configuration>
              <executions>
                <execution>
                  <id>make-assembly</id>
                  <phase>package</phase>
                  <goals>
                    <goal>single</goal>
                  </goals>
                </execution>
              </executions>
            </plugin>
          </plugins>
        </build>

二、关于scala的配置

setting中java compiler-> 1.8 ;project structure -> 8; libraries 中添加scala-sdk

三、目录

其中日志文件：log4j.properties

#定义输出级别：只输出 DEBUG类型，输出到console和File
#log4j.rootLogger=debug,Console,File
log4j.rootLogger=WARN,File

#日志输出方式：控制台输出
#log4j.appender.Console=org.apache.log4j.ConsoleAppender
#log4j.appender.Console.Target=System.out

#可以灵活地指定布局模式
log4j.appender.Console.layout=org.apache.log4j.PatternLayout
log4j.appender.Console.layout.ConversionPattern = %d{yyyy-MM-dd HH:mm} -%p (%F:%L)- %m%n

### 配置输出到文件，并且每天都创建一个文件 ###
log4j.appender.File = org.apache.log4j.DailyRollingFileAppender
### windows下输出路径：全路径 d:/logs/myspark.log
### linux下输出路径(当前路径下的)：logs/myspark.log
log4j.appender.File.File = d:/logs/myspark.log
#### DatePattern 确定创建文件的周期默认为每天 ####
log4j.appender.File.DatePattern='_' yyyy-MM-dd
log4j.appender.File.Append = true
log4j.appender.File.Threshold = DEBUG
log4j.appender.File.layout = org.apache.log4j.PatternLayout
log4j.appender.File.layout.ConversionPattern = %-d{yyyy-MM-dd HH:mm:ss} [ %t:%r ] - [ %p ] %m%n

四、spark代码

（一）介绍

1）spark RDD

    // 1, sparkConf
    val conf = new SparkConf().setMaster("local[2]").setAppName("myjob")// 本地2核 ；*表示所有的核
    // 2, sparkcontext
    val sc = new SparkContext(conf) //创建SparkContext，该对象是提交spark App的入口
    // 3, 使用sc创建RDD并执行相应的transformation和action
        sc.textFile("file:////root/sparkdata/wordcount.txt")
      //  sc.textFile("file:////D:\idea\ideaProjects\spark_projects\myspark8\src\main\scala\com\njbdqn\wordcount.txt")
          .flatMap(_.split(" ")).map((_,1)).reduceByKey(_+_).foreach(println(_))
    // 4, 停止sc，结束该任务
        sc.stop()

2）spark sql

    val spark = SparkSession.builder().master("local[*]").appName("tb").getOrCreate()
    // 读取hdfs上的文件
    val df = spark.read.format("csv").load("hdfs://192.168.56.111:9000/quizz/data/userbehaviour/UserBehavior.csv")
    df.show(5)
    spark.stop()

（二）读各种格式的文件

① 变成DataFrame

1）CSV
读csv的sc方法

val rdd1 = sc.textFile("hdfs://192.168.56.111:9000/party/data/locale.csv")
val df1 = rdd1.map(x=>{
val strs = x.split(",")
(strs(0),strs(1))
}).toDF("id","locale")

读csv的sparkssession方法

val spark = SparkSession.builder().master("local[*]").appName("tb").getOrCreate()
// 读取hdfs上的文件
val df = spark.read.format("csv").option("header","true").load("hdfs://192.168.56.111:9000/quizz/data/userbehaviour/UserBehavior.csv")
df.show(5)
spark.stop()

2）TXT
【常用】读txt使用sc

val rdd1 = sc.textFile("hdfs://192.168.56.111:9000/party/data/locale.txt")
val df1 = rdd1.map(x=>{
val strs = x.split("	")
(strs(0),strs(1))
}).toDF("id","locale")

读txt使用sparksession：

spark.read.option("sep","	").csv("路径")
.withColumnRenamed("_c0","name1")
.withColumnRenamed("_c1","name2")
.show

② 用rdd做

用sc读取csv文件，转成RDD[Array[String]]，并去表头

val rddpre = sc.textFile("hdfs://192.168.56.111:9000/party/data/event_attendees.csv").map(x=>x.split(",",-1))
val head = rddpre.first()
val rdd = rddpre.filter(x=>x!=head).cache

 str.split(",",-1) 和 str.split(",") 的区别：前者把","之间的空值保留，后者去除空值

由于是csv文件，如果把1,2,3,,5变成1,2,3,5 则二维的对应列就乱了

五、打包

maven打包package后，把target下胖包（是with-dependences那个jar包）放到linux下java -jar 路径 执行

 ./spark-submit --class com.njbdqn.MySpark /root/myspark-jar-with-dependencies.jar