大数据（7n）Scala隐式转换

Posted 2021-05-26 小基基o_O

低调隐帅的目录

• 隐式函数
• 隐式类
• 隐式参数
• 写Spark时用到隐式转换

隐式函数

import scala.language.implicitConversions

object Hello {
  def main(args: Array[String]): Unit = {
    val i: Int = 1.0
    print(i) // 打印结果：1
  }

  implicit def double2int(d: Double): Int = d.toInt
}

如上代码所示，关键词implicit使Double可以转成Int

隐式类

import scala.language.implicitConversions

object Hello {
  def main(args: Array[String]): Unit = {
    val a: A = new A("aa")
    a.methodA // 打印结果：Aaa
    a.methodB // 打印结果：BAaa
  }

  class A(s: String) {
    val data: String = "A" + s
    val methodA: Unit = println(data)
  }

  class B(s: String) {
    val data: String = "B" + s
    val methodB: Unit = println(data)
  }

  implicit def a2b(a: A): B = new B(a.data)
}

import scala.language.implicitConversions

object Hell0 {
  def main(args: Array[String]): Unit = {
    val a: A = new A("aa")
    a.methodA
    a.methodB
  }

  class A(s: String) {
    val data: String = "A" + s
    val methodA: Unit = println(data)
  }

  implicit class B(a: A) {
    val data: String = "B" + a.data
    val methodB: Unit = println(data)
  }
}

如上两个代码，第1种是隐式函数，第2种从第1种变过来，是隐式类

隐式参数

// 方法
def m(k: Int, b: Int)(implicit x: Int) = k * x + b

// 传参
val y1 = m(2, -2)(1)
println(y1)

// 隐式参数：自动传参
implicit val x: Int = 1
val y2 = m(2, -2)
print(y2)

写Spark时用到隐式转换

RDD、DataFrame、DataSet 3者互转

val c1: SparkConf = new SparkConf().setAppName("a1").setMaster("local[*]")
val spark: SparkSession = SparkSession.builder().config(c1).getOrCreate()
//隐式转换支持
import spark.implicits._