Skip to content

mtumilowicz/kotlin-dsl-lambda-workshop

Repository files navigation

Build Status License: GPL v3

kotlin-dsl-lambda-workshop

preface

  • goals of this workshop:
    • introduce fundamental kotlin concepts
      • extension function
      • lambda with receiver
      • infix operator
      • invoke operator
      • overloading operators
      • object destructuring
    • understanding what is state machine and how it works
    • becoming acquainted with DSL and how kotlin supports it
  • workshop: workshop package, answers: answers package

extension function

  • top-level function
    • you can place functions directly at the top level of a source file, outside of any class
    • when you compile the file, some classes will be produced
      • JVM can only execute code in classes
    • example
      package strings // filename: StringUtils.kt
      fun isNotEmpty(...): Boolean { ... }
      
      /* Java */
      package strings;
      public class StringUtilsKt { // corresponds to the filename
          public static Boolean isNotEmpty(...) { ... }
      }
      
  • we can define them inside classes as members - therefore narrow the scope
  • extension function it’s a function that can be called as a member of a class but is defined outside of it
    • example
      fun String.lastChar(): Char = this.get(this.length - 1)
      fun String.lastChar(): Char = get(length - 1) // this is implicit
      
      • receiver type: String - type on which the extension is defined
      • receiver object: this - the instance of that type

lambdas with receivers

  • the ability to call methods of a different object in the body of a lambda
  • example (from standard library - look for others)
    • apply - calls the specified function block with this value as its receiver and returns this value
      @kotlin.internal.InlineOnly
      public inline fun <T> T.apply(block: T.() -> Unit): T {
          contract {
              callsInPlace(block, InvocationKind.EXACTLY_ONCE)
          }
          block() // this.block()
          return this
      }
      
    • use case
      val name = StringBuilder().apply {
          append("M")
          append("i")
          append("chal")
          toString()
      }
      println(name) // Michal
      
  • lambda with a receiver looks exactly the same as a regular lambda in the source code
    fun createClient(c: ClientBuilder.() -> Unit): Client {
        val builder = ClientBuilder()
        c(builder) // or builder.c()
        return builder.build()
    }
    
    val client = createClient {
        firstName = "Anton" // implicit this
        lastName = "Arhipov"
    
        twitter { // method call: this.twitter(...)
            name = "@antonarhipov"
        }
    }
    
    • is same as
    fun createClient(c: (ClientBuilder) -> Unit): Client { // difference here
        val builder = ClientBuilder()
        c(builder)
        return builder.build()
    }
    
    val client = createClient {
        it.firstName = "Anton" // explicit it
        it.lastName = "Arhipov"
        it.twitter(...)
    }
    
  • lambdas with receiver and extension functions
    • note that an extension function is, in a sense, a function with a receiver
      • this refers to the instance of the type the function is extending
    • analogy
      • regular function <-> regular lambda
      • extension function <-> lambda with a receiver
    • method-name conflicts
      • use this@OuterClass.conflictedMethod()
    • what is exactly T.()?
      • fun doSomethingWithInt(receiver: Int.() -> Unit)
      • how this is different from Regular lambda like () -> Unit?
      • similar to extension function we could call: 5.receiver()
      • you can also write receiver(5)

infix

  • example
    • 1 to "one" is same as 1.to("one")
  • the method name is placed between the target object and the parameter, with no extra separators
  • to unlock infixing, you need to mark function with the infix modifier

invoke operator

  • we could invoke function types (single abstract method: invoke) in a very simple way
    val function: (String) -> Boolean = {...}
    val argument = "Michal"
    function(argument) // internally calls SAM
    
  • invoke is an operator
  • example
    class Issues(
        val data: MutableList<String> = mutableListOf()
    ) {
        fun add(s: String) {
            data.add(s)
        }
    
        operator fun invoke(body: Issues.() -> Unit) {
            body()
        }
    }
    
    adding feature to call lambda outside of brackets we get
    val issues = Issues()
    issues.add("a")
    
    issues { // issues({add("abc")}) -> issues() {add("abc")} -> issues {add("abc")}
        add("abc")
    }
    

destructuring

  • example
    val p = Point(10, 20)
    val (x, y) = p  
    
    for ((key, value) in map) {
        println("$key -> $value")
    }
    
  • to initialize each variable - a function named componentN is called
    • N is the position of the variable in the declaration
  • for a data class, the compiler generates a componentN function for every property declared in the primary constructor
    • vanilla example
      class Point(val x: Int, val y: Int) {
          operator fun component1() = x
          operator fun component2() = y
      }      
      
  • use-case
    val (name, ext) = "filename.exe".split(".", limit = 2)
    println("$name $ext") // filename exe
    
    • destructuring could throw exceptions in case of too few arguments
      • Exception in thread "main" java.lang.IndexOutOfBoundsException: Index: 1, Size: 1

operator overloading

dsl

  • there’s no well-defined boundary between a DSL and a regular API
    • one trait especially characteristic for DSL - structure (grammar)
  • in API - no inherent structure in the sequence of calls, no context maintained between one call and the next
    • sometimes called command-query API
  • the method calls in a DSL exist in a larger structure, defined by the grammar of the DSL
    • In a Kotlin - structure is created by nesting lambdas and chained method calls
  • grammar is what allows us to call an internal DSL a language
    • similar to a natural language such as English
      • the function names usually act as verbs ( groupBy , orderBy )
      • their arguments fulfill the role of nouns ( Country.name )
      • single operation (tense) can be composed out of multiple function calls (words)
      • the type checker ensured the calls are combined in a meaningful way
  • kotlin support for DSL
    • extension function
    • infix call
    • operator overloading
      • 2 in listOf(1, 2) // listOf(1,2).contains(2)
      • map["key"] // map.get("key")
    • lambda outside of parentheses
      • file.use { it.read() } // file.use({ f -> f.read() } )
    • lambda with a receiver