@[toc]

入门

概述

https://www.scala-lang.org/

Scala combines object-oriented and functional programming in one concise, high-level language. Scala's static types help avoid bugs in complex applications, and its JVM and JavaScript runtimes let you build high-performance systems with easy access to huge ecosystems of libraries.

• 学习Scala的意义：
  Spark、Kafka、Flink
  优雅
  开发速度快
  融合到生态圈

安装

1. 安装 Java 8
2. 下载 download scala 网址：https://www.scala-lang.org/download/2.11.8.html
3. 解压 unzip scala
4. 配置环境变量(可选)

Windows 需配置两个 Path中： D:\scala\bin 和 D:\scala\jre\bin

1. 查看是否生效

Linux或Mac中操作步骤：
1.tar -zxvf scala-2.11.8.tgz -C  解压路径
2.到解压目录下 pwd  复制整个路径 
3.将上面的路径 添加到环境变量中
    vi ~/.bash_profile
        export SCALA_HOME=复制的路径
        export PATH=$SCALA_HOME/bin:$Path
    保存
    source ~/.bash_profile
    echo $SCALA_HOME

下载之后的scala目录下的bin目录中 有普通文件 和 .bat文件
.bat文件是在Windows中用的，Linux或Mac中用不到，所以可以删掉 rm *.bat

Java VS Scala

Java

public class HelloWorld {
    public static void main(String[] args) {
        System.out.println("Hello World..");
    }
}

Scala(每行代码并不强求使用;结束，但是Java是必须的)

Object HelloWorld {
    def main(args : Array[String]) {
        println("Hello World..")
    }
}

val 和 var

• val：值

final
val 值名称：类型=xx
val a = 1  (不可变)
val a : int = 1

• var：变量

var 值名称：类型=xxx
var b = 1
var b : int = 1

基本数据类型

• Byte/Char
• Short/Int/Long/Float/Double
• Boolean

只有Float声明时比较特别

• var c : Float = 1.1f

scala> b=20
b: Int = 20

scala> val b:Int =10
b: Int = 10

scala> val c:Boolean=true
c: Boolean = true

scala> val d =1.1
d: Double = 1.1

scala> val e:Float=1.2f
e: Float = 1.2

lazy在Scala中的应用

lazy var d : int = 1;
延迟加载，只有在第一次使用时才加载

读取文件并以字符串形式输出

    import scala.io.Source._
    var info = fromFile("...").mkString

如果用lazy var info = fromFile("...").mkString，开始是检测不到错误的，要小心使用

*注意：当一个变量声明为lazy，只有当你第一次操作时才会去真正访问，如果不去访问，即使写错了，也不会发现

开发工具IDEA

Maven

1.下载IDEA和Maven
2.进入IDEA，新建项目 选择Maven 勾选create from archetype 选择scala-archetype simple-> 正常创建(注意Maven仓库位置)
3.IDEA默认是不支持Scala的，需要下载Scala插件
    File -> settings -> Plugins -> install JetBrains plugin -> scala
    之后就可以new 一个Scala类了
4.新建测试类，运行报错
    删除pom.xml中<arg>make:transitive</args>

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
  <modelVersion>4.0.0</modelVersion>
  <groupId>org.example</groupId>
  <artifactId>untitled5</artifactId>
  <version>1.0-SNAPSHOT</version>
  <inceptionYear>2008</inceptionYear>
  <properties>
    <maven.compiler.source>1.8</maven.compiler.source>
    <maven.compiler.target>1.8</maven.compiler.target>
    <encoding>UTF-8</encoding>
    <scala.version>2.11.8</scala.version>
    <spark.version>2.3.0</spark.version>
  </properties>

  <dependencies>
    <dependency>
      <groupId>org.scala-lang</groupId>
      <artifactId>scala-library</artifactId>
      <version>${scala.version}</version>
    </dependency>

    <!--引入Spark Core的依赖-->
    <!-- https://mvnrepository.com/artifact/org.apache.spark/spark-core -->
    <dependency>
      <groupId>org.apache.spark</groupId>
      <artifactId>spark-core_2.11</artifactId>
      <version>2.3.0</version>
    </dependency>

    <!-- https://mvnrepository.com/artifact/org.apache.spark/spark-sql -->
    <dependency>
      <groupId>org.apache.spark</groupId>
      <artifactId>spark-sql_2.12</artifactId>
      <version>2.4.3</version>
    </dependency>

  </dependencies>

  <build>
    <sourceDirectory>src/main/scala</sourceDirectory>
    <testSourceDirectory>src/test/scala</testSourceDirectory>
    <plugins>
      <plugin>
        <groupId>org.scala-tools</groupId>
        <artifactId>maven-scala-plugin</artifactId>
        <version>2.15.0</version>
        <executions>
          <execution>
            <goals>
              <goal>compile</goal>
              <goal>testCompile</goal>
            </goals>
            <configuration>
              <args>
                <arg>-dependencyfile</arg>
                <arg>${project.build.directory}/.scala_dependencies</arg>
              </args>
            </configuration>
          </execution>
        </executions>
      </plugin>
      <plugin>
        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-surefire-plugin</artifactId>
        <version>2.6</version>
        <configuration>
          <useFile>false</useFile>
          <disableXmlReport>true</disableXmlReport>
          <!-- If you have classpath issue like NoDefClassError,... -->
          <!-- useManifestOnlyJar>false</useManifestOnlyJar -->
          <includes>
            <include>**/*Test.*</include>
            <include>**/*Suite.*</include>
          </includes>
        </configuration>
      </plugin>
    </plugins>
  </build>
</project>

函数

方法定义

    def 方法名(参数: 参数类型): 返回值类型 = {
        //方法体

        //最后一行作为返回值(不需要使用return)
    }

    def max(x: Int, y: Int): Int = {
        if(x > y)
            x
        else 
            y
    }

package org.example

object App {

  def main(args: Array[String]): Unit = {
    println(add(2,5))
  }


  def add(x:Int,y:Int):Int={
  x+y
  }
}

7

package org.example

object App {

  def main(args: Array[String]): Unit = {

    println(three())
    //没有入参的时候可以不用写
    println(three)
  }




  def three()=1+2
}

无返回值 自动加Unit

默认参数

默认参数： 在函数定义时，允许指定参数的默认值

  //参数
  def sayName(name: String ) = {
    println(name)
  }

  //默认参数
  def sayName1(name: String ="Jack") = {
    println(name)
  }


//main调用

    sayName("jaja")
    sayName1()
    sayName1("Ma")

jaja
Jack
Ma

相关源码：SparkContext中使用

命名参数

可以修改参数的传入 顺序

      def speed(destination: Float, time: Float): Float {
          destination / time
      }

      println(speed(100, 10))
      println(speed(time = 10, destination = 100))

可变参数

变参数(可传入任意多个相同类型的参数) java中 int... numbers

JDK5+：可变参数

    def sum(number: Int*) =  {
        var result = 0
        for(num <- number) {
            result += num
        }
        result
    }

相关源码：org.apache.spark.sql.Dataset中的select方法

条件语句

循环语句

• to 1 to 10 (左闭右闭) 1.to(10)
• range Range(1,10) (左闭右开的) Range(1,10,2) (2为步长)
• until 1 until 10 (左闭右开)

to、until的底层调用都是Range

scala> 1 to 10
res1: scala.collection.immutable.Range.Inclusive = Range(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)

scala> Range(1,10)
res2: scala.collection.immutable.Range = Range(1, 2, 3, 4, 5, 6, 7, 8, 9)

scala> 1.to(10)
res3: scala.collection.immutable.Range.Inclusive = Range(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)

scala> Range(1,10,2)
res4: scala.collection.immutable.Range = Range(1, 3, 5, 7, 9)

scala> Range(1,10,5)
res5: scala.collection.immutable.Range = Range(1, 6)

scala> Range(10,1,-1)
res8: scala.collection.immutable.Range = Range(10, 9, 8, 7, 6, 5, 4, 3, 2)

scala> 1 until 10
res9: scala.collection.immutable.Range = Range(1, 2, 3, 4, 5, 6, 7, 8, 9)

• for

    for(i <- 1.to(10)) {
        println(i)
    }
    for(i <- 1.until(10, 2)) {
        println(i)
    }
    for(i <- 1 to 10 if i % 2 == 0) {
        println(i)
    }

    val courses = Array("Hadoop", "Spark SQL", "Spark Streaming", "Storm", "Scala")
    for(x<- courses) {
        println(x)
    }

    //x其实就是courses里面的每个元素
    // =>      就是将左边的x作用上一个函数，变成另外一个结果


    courses.foreach(x=> println(x))

• while

    var (num, sum) = (100, 0)
    while(num > 0){
        sum = sum + num
        num = num - 1
    }
    println(sum)

面向对象

概述

• Java/Scala OO(Object Oriented)

  • 封装：属性、方法封装到类中，可设置访问级别
  • 继承：父类和子类之间的关系 重写
  • 多态：父类引用指向子类对象 开发框架基石

Person person = new Person();
User user = new User();
Person person =new User();

类的定义和使用

package org.example

object ObjectApp {
  def main(args: Array[String]): Unit = {
    val person = new People()
    person.name = "Messi"
    //    println(person.name + ".." + person.age)
    println("invoke eat method: " + person.eat)
    person.watchFootball("Barcelona")

    person.printInfo()
    //编译不通过 private 修饰
    //    println(person.gender)
  }

}



class People{
  //var(变量)类型自动生成getter/setter
  //这种写法就是一个占位符
  var name: String = _
  //val(常量)类型自动生成getter
  val age: Int = 10

  private [this] var gender = "male"

  def printInfo() : Unit = {
    print("gender: " + gender)
  }

  def eat(): String = {
    name + " eat..."
  }

  def watchFootball(teamName: String): Unit = {
    println(name + " is watching match of " + teamName)
  }
}

invoke eat method: Messi eat...
Messi is watching match of Barcelona
gender: male

继承和重写

• 继承

class Student(name: String, age: Int, var major: String) extends Person(name, age) {}

• 重写

    override def acquireUnrollMemory()
    override def toString = "test override"

package org.example

object ConstructorApp {
  def main(args: Array[String]): Unit = {
    var person =new Person("zhangsan",99)
    println(person.age+":"+person.name)
    var person2 =new Person("zhangsan",99,"Man")
    println(person2.age+":"+person2.name+";"+person2.gender)
  }

}
//主构造器
class Person(val name: String, val age: Int){
  println("Person constructor enter...")

  val school = "ustc"
   //占位符肯定要预先指定类型
  var gender: String = _

  //附属构造器
  def this(name: String , age: Int, gender: String){
    //必须要调用主构造器或者其他附属构造器
    this(name, age)
    this.gender = gender
  }

  override def toString = "test override"
  println("Person Constructor leave...")

}

//继承
//name: String, age: Int, var major: String 继承父类的可以不用直接写var 否则需要重新申明
class Student(name: String, age: Int, var major: String) extends Person(name, age) {

  //重写
  override val school = "pku"
  println("Person Student enter...")

  println("Person Student leave...")
}

抽象类

package org.example

object AbstractApp {
  def main(args: Array[String]): Unit = {
    var stu =new Student1();
    println(stu.age)
    println(stu.name)
    stu.speak;
  }
}
abstract class Person3{
  def speak

  val name: String
  val age: Int
}

class Student1 extends Person3{
  override def speak: Unit = {
    println("speak")
  }
  override val name: String = "Messi"
  override val age: Int = 32
}

伴生类和伴生对象

如果有一个 class，还有一个与 class同名的 object 互为 伴生类和伴生对象

    class ApplyTest{
        def apply(){
            println(...)
        }
    } 
    object ApplyTest{
        def apply(){
            println("Object ApplyTest apply...")
            new ApplyTest
        }
    } 

类名() ==> Object.apply

  对象() ==> Class.apply  
  最佳实践：在Object的apply方法中去new一个Class

package org.example


object ApplyApp {
  def main(args: Array[String]): Unit = {
//    for(i<-1 to 10){
//      ApplyTest.incr
//    }
//    //object 是一个单例对象
//    println(ApplyTest.count)

    var b=ApplyTest()
    //默认走的是object=》apply
    //Object ApplyTest apply...

    println("-----------------------")
    var c= new ApplyTest()
    c()
    //Class ApplyTest apply...
  }

}
class ApplyTest {
  def apply() = {
    println("Class ApplyTest apply...")

  }

}
object ApplyTest {
  println("Object start...")
  var count = 0
  def incr={
    count=count+1
  }
  def apply() = {
    println("Object ApplyTest apply...")

    //在object中的apply中new class
    new ApplyTest
  }
  println("Object end...")
}

case和trait

case class ：不用new

case class Dog(name: String)
直接可以调用Dog("wangcai")

Trait： 类似implements

xxx entends ATrait 
xxx extends Cloneable with Logging with Serializable

源码中Partition类

集合

数组

package org.example

object ArrayApp extends App{
  //println("hello")
  val a = new Array[String](5)
  a(0)="hello"
  println(a(0))

  val b = Array("hello","world")

  val c = Array(1,2,3,4,5,67)
  c.sum
  c.max
  c.mkString("/")

}

  val d=scala.collection.mutable.ArrayBuffer[Int]()

  d+=1
  d+=2
  d+=(2,33,4)
  d++=Array(33,45,22)
  println(d+"-------------------")
  d.insert(0,999)
  d.remove(1,2)
  d.trimEnd(2)
  println(d+"-------------------")
  //转化成不可变的
  d.toString()
  for(i<-0 until d.length){

    println(c(i))
  }

hello
ArrayBuffer(1, 2, 2, 33, 4, 33, 45, 22)-------------------
ArrayBuffer(999, 2, 33, 4, 33)-------------------
1
2
3
4
5

List

list是不可变的，对list进行添加删除或者取值等操作均会返回一个新的list。

scala> Nil
res4: scala.collection.immutable.Nil.type = List()

scala> Nil
res4: scala.collection.immutable.Nil.type = List()

scala> val l= List(1,2,3,4,5,56)
l: List[Int] = List(1, 2, 3, 4, 5, 56)

scala> l.head
head   headOption

scala> l.head
res5: Int = 1

scala> l.tail
tail   tails

scala> l.tail
res6: List[Int] = List(2, 3, 4, 5, 56)

scala> l.tails
res7: Iterator[List[Int]] = non-empty iterator

scala>

 val d=scala.collection.mutable.ArrayBuffer[Int]()

  d+=1
  d+=2
  d+=(2,33,4)
  d++=Array(33,45,22)
  d++ =List(1,2,3,4,)

Set

set是一个非重复的集合，若有重复数据，则会自动去重。

scala> val set = Set(1,2,3,1,2,5)
set: scala.collection.immutable.Set[Int] = Set(1, 2, 3, 5)

Map

map是K-V键值对集合。

package org.example

object MapApp {

  def main(args: Array[String]): Unit = {

    val map = Map(
      "1" -> "hello" ,
      2 -> "world",
      3 -> "!!!!!"
    )
    println(map.mkString(","))

    println("-----------------------")
    for(x<-map){
      println(x._1+":"+x._2)
    }

    println("-----------------------")

    var keys = map.keys
    var keyIterator = keys.iterator
    while(keyIterator.hasNext) {
      val key = keyIterator.next()
      println(key + "\t" + map.get(key).get)
    }



  }

}

 1 -> hello,2 -> world,3 -> !!!!!
-----------------------
1:hello
2:world
3:!!!!!
-----------------------
1    hello
2    world
3    !!!!!

Optuon&Some&None

    val map = Map(
      "1" -> "hello" ,
      2 -> "world",
      3 -> "!!!!!"
    )

    println(map.get(2))
    println(map.get(999))

Some(world)
None

option.scala

@SerialVersionUID(5066590221178148012L) // value computed by serialver for 2.11.2, annotation added in 2.11.4
case object None extends Option[Nothing] {
  def isEmpty = true
  def get = throw new NoSuchElementException("None.get")
}


@SerialVersionUID(1234815782226070388L) // value computed by serialver for 2.11.2, annotation added in 2.11.4
final case class Some[+A](x: A) extends Option[A] {
  def isEmpty = false
  def get = x
}

Tuple

与列表一样，与列表不同的是元组可以包含不同类型的元素。元组的值是通过将单个的值包含在圆括号中构成的。创建过程可加new关键词，也可不加。

package org.example

object TupleApp {

  def main(args: Array[String]): Unit = {

    var t=new Tuple3[Int,Int,String](1,99,"hello")
    println(t.toString())

    println("----------------")

    var t2=(9999,"hello")

    println(t2.toString())

    println(t2.swap.toString())


  }

}

(1,99,hello)
----------------
(9999,hello)
(hello,9999)

模式匹配

基本类型

Java ： 对一个值进行条件判断例如switch

模式匹配类似于java的switch case。Scala的模式匹配不仅可以匹配值还可以匹配类型、从上到下顺序匹配，如果匹配到则不再往下匹配、都匹配不上时，会匹配到case _ ,相当于default、match 的最外面的”{ }”可以去掉看成一个语句。

def match_test(m:Any) = {
       m match {
         case 1 => println("nihao")
         case m:Int => println("Int")
         case _ => println("default")
       }
    }

package org.example

object MarchApp {

  def main(args: Array[String]): Unit = {

    def judeGrade(grade:String)={
      grade match{
        case "B" => println("Just so so")
        case "A" => println("good")
        case "S" => println("cool")
        case _=> println("No.1")
      }
      
    }
    judeGrade("S")
    judeGrade("A")
    judeGrade("SSS")
  }

}

cool
good
No.1

Array

    def greeting(array:Array[String]) = {
      array match {
        case Array("zs")=> println("hi,zs")
        case Array(x,y)=> println(x+"and"+y)
        case Array("zs",_*)=>println("zs and other")
        case _=>println("everyone")
      }
    }

    greeting(Array("zs"))
    greeting(Array("zs","ls"))

hi,zs
zsandls

List

    def greeting1(list:List[String]) = {
      list match {
        case "zs"::Nil=> println("hi,zs")
        case x::y::Nil=> println(x+"and"+y)
        case "zs"::tail =>println("zs and other")
        case _=>println("everyone")
      }
    }

    greeting1(List("zs"))
    greeting1(List("zs","ls"))
  }

类型匹配

    def matchType(obj: Any) = {
      obj match {
        case x: Int => println("hi,int")
        case y: String => println(y)
        case m: Map[_, _] => println("map")
        case _ => println("everyone")
      }
    }

    matchType(Map(1 -> "yes"))
    matchType(11)
    matchType("hello")
  }

异常处理

 try{
      val i=10/0
      println(i)

    }catch {
      case e:ArithmeticException=>println(e.getMessage)
      case e:Exception=>println(e.getMessage)
    }finally {

    }

/ by zero

高级函数

字符串

• 插值

  val s ="hello"
  val name="jacksun"
  println(s+name)
  println(s+":"+name)

  println(s"hello:$name")

• 多行字符串

  //多行
  var d =
    """
      |1
      |2
      |3
      |4
      |5
      |5
      |6
    """.stripMargin

匿名函数

匿名函数分为有参匿名函数、无参匿名函数、有返回值的匿名函数。（可以将匿名参数的返回给一个val声明的值，匿名函数不能显式的声明返回值）

package org.example

object FunctionApp extends App {

  //有参数匿名函数
  val printy = (a : Int) => {
    println(a)
  }
  printy(999)

  //无参数匿名函数
  val printx = ()=>{
    println("Scala No.1")
  }
  printx()

  //有返回值的匿名函数
  val add = (a:Int,b:Int) =>{
    a+b
  }
  println(add(4,4))

}

Currying

将接受一个参数的转化成2个

def add(a:Int,b:Int) = a+b
println(add(2,1))
//Currying
def add2(a:Int)(b:Int) = a+b
println(add2(2)(1))

高阶函数

高阶函数（Higher-Order Function）就是操作其他函数的函数。

Scala 中允许使用高阶函数, 高阶函数可以使用其他函数作为参数，或者使用函数作为输出结果。

object Test {
   def main(args: Array[String]) {

      println( apply( layout, 10) )

   }
   // 函数 f 和 值 v 作为参数，而函数 f 又调用了参数 v
   def apply(f: Int => String, v: Int) = f(v)

   def layout[A](x: A) = "[" + x.toString() + "]"
   
}

• map

对每个集合的元组进行操作

scala> val l =List(1,2,3,4,5,6,7,8,9)
l: List[Int] = List(1, 2, 3, 4, 5, 6, 7, 8, 9)

scala> l.map(x=>(x+1))
res5: List[Int] = List(2, 3, 4, 5, 6, 7, 8, 9, 10)

scala> l.map((x:Int)=>x*2)
res6: List[Int] = List(2, 4, 6, 8, 10, 12, 14, 16, 18)

scala> l.map(x=>x*2)
res7: List[Int] = List(2, 4, 6, 8, 10, 12, 14, 16, 18)

scala> l.map(_*2)

• filter

过滤条件

scala> l.filter(_>5)
res9: List[Int] = List(6, 7, 8, 9)

• take

取数

scala> l.take(1)
res10: List[Int] = List(1)

scala> l.take(3)
res11: List[Int] = List(1, 2, 3)

• reduce

两两相加相减

scala> l.take(3).reduce(_-_)
res15: Int = -4
// 从左相减
scala> l.take(3).reduceLeft(_-_)
res16: Int = -4
// 从右相减
scala> l.take(3).reduceRight(_-_)
res17: Int = 2