SparkSQL自定义UDF函数

￡神魔★判官ぃ 2024-02-18 10:32 67阅读 0赞

需求：员工id正常为8位，对于不满8位的员工id左侧用0补齐

import org.apache.spark.sql.{
        DataFrame, SparkSession}
    
    object DataSetCreate {
        
    
      def main(args: Array[String]): Unit = {
        
        val spark = SparkSession
          .builder()
          .appName("test")
          .master("local[4]")
          .getOrCreate()
        import spark.implicits._
    
        val list = List(
          ("1001","zhangsan",10),
          ("44","zhang2",66),
          ("32432","zhang3",70),
          ("6757","zhang4",22),
          ("1","zhang4",22))
        val df: DataFrame = list.toDF("id","name","age")
    
        //需求：员工id正常为8位，对于不满8位的员工id左侧用0补齐
        //println(myLpad("h",8,"0"))
        spark.udf.register("xxx",myLpad _);
        df.selectExpr("xxx(id,8,'0') id","name","age").show()
    
        df.createOrReplaceTempView("data")
        spark.sql(
          """
            |select
            |xxx(id,8,'0') id,
            |name,
            |age
            |from data
            |""".stripMargin).show()
    
      }
    
        /**
         *
         * @param str 待补齐的字符串
         * @param len 最终补齐的长度
         * @param pad 使用pad填充
         * @return
         */
        def myLpad(str:String,len:Int,pad:String):String={
        
          //字符串长度
          val strLen: Int = str.length
          //需要补齐的长度
          val padLen: Int = len-strLen
          s"${pad * padLen}${str}"
        }
    
    }

![在这里插入图片描述][126b40b4712d42dc8537419fe202c345.png]

[126b40b4712d42dc8537419fe202c345.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/02/18/485d0e5b9b304f94bb12d9eb34932d2a.png