Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Scala — An Introduction

Scala — An Introduction

An introduction to Scala tailored to Java developers. Presentation given at the Bucharest Java User Group.

Ionuț G. Stan

July 25, 2013
Tweet

More Decks by Ionuț G. Stan

Other Decks in Programming

Transcript

  1. Ionuț G. Stan · [email protected] · igstan.ro Agenda ‣ What’s

    Scala ‣ Why Scala ‣ Quick language overview
  2. Ionuț G. Stan · [email protected] · igstan.ro Martin Odersky ‣

    Scala’s designer ‣ Helped design Java’s generics (Generic Java) ‣ Worked on the current javac compiler
  3. Ionuț G. Stan · [email protected] · igstan.ro Martin Odersky Generic

    Java team (left to right): Philip Wadler, Martin Odersky, Gilad Bracha, David Stoutamire.
  4. Ionuț G. Stan · [email protected] · igstan.ro Martin Odersky Generic

    Java team (left to right): Philip Wadler, Martin Odersky, Gilad Bracha, David Stoutamire.
  5. Ionuț G. Stan · [email protected] · igstan.ro What is Scala

    ‣ A language for the JVM ‣ Statically typed ‣ Compiles to JVM bytecode, just like Java ‣ Research on compiling to other platforms ‣ The JVM has influenced Scala’s semantics ‣ Language + compiler + standard library
  6. Ionuț G. Stan · [email protected] · igstan.ro What is Scala

    ‣ Supports two programming paradigms ‣ Object-Oriented Programming (OOP) ‣ Functional Programming (FP) ‣ Created to explore this mix ‣ Martin Odersky sees FP and OOP as orthogonal ‣ OOP does not imply imperative
  7. Ionuț G. Stan · [email protected] · igstan.ro What is FP

    ‣ Various opinions ‣ All centered around the same concept ‣ FP is programming w/ functions/values ‣ FP is programming w/o side-effects/mutations ‣ FP’s roots are in mathematical functions ‣ A function’s output is determined entirely by its input
  8. Ionuț G. Stan · [email protected] · igstan.ro Why FP ‣

    Easier to go concurrent/parallel ‣ Easier to reason about code ‣ Fosters composability
  9. Ionuț G. Stan · [email protected] · igstan.ro Scala and FP

    ‣ Encourages immutability (syntax + stdlib) ‣ Concise syntax for function literals ‣ Function types ‣ Pattern matching ‣ Case classes ‣ Favours expressions over statements ‣ Lazy evaluation
  10. Ionuț G. Stan · [email protected] · igstan.ro Why Scala ‣

    Improves support for OOP ‣ Good support for FP ‣ Runs on JVM, one of the best VMs around ‣ A better Java ‣ Yet interoperable with Java ‣ Powerful type system (vs. Clojure) ‣ Favours immutability, but allows mutability
  11. Ionuț G. Stan · [email protected] · igstan.ro Language Features ‣

    Type inference ‣ Function types and literals ‣ Objects as functions ‣ Case classes ‣ Pattern matching ‣ Traits ‣ Lazy evaluation ‣ Macros (compile-time metaprogramming)
  12. Ionuț G. Stan · [email protected] · igstan.ro Hello World package

    ro.bjug.fourteen object Main { def main(args: Array[String]): Unit = { val name: String = args(0) println(s"Hello $name!") } }
  13. Ionuț G. Stan · [email protected] · igstan.ro Type Inference package

    ro.bjug.fourteen object Main { def main(args: Array[String]) = { val name = args(0) println(s"Hello $name!") } }
  14. Ionuț G. Stan · [email protected] · igstan.ro Objects as Functions

    object Factorial { def apply(n: Int): Int = { if (n < 2) { return 1 } else { return n * Factorial(n - 1) } } } Factorial(5)
  15. Ionuț G. Stan · [email protected] · igstan.ro Expressions object Factorial

    { def apply(n: Int): Int = { if (n < 2) 1 else n * Factorial(n - 1) } } Factorial(5)
  16. Ionuț G. Stan · [email protected] · igstan.ro Function Literals val

    double: Int => Int = (n) => n * 2 val double = (n: Int) => n * 2 double(5)
  17. Ionuț G. Stan · [email protected] · igstan.ro Functions as Objects

    val double = (n: Int) => n * 2 val square = (n: Int) => n * n double(square(2)) // equivalent to square.andThen(double)(2)
  18. Ionuț G. Stan · [email protected] · igstan.ro Classes class Person(name:

    String, age: Int) { def bio: String = s"$name is a $age years old developer." } val person = new Person("Ionuț", 28) person.bio // Ionuț is a 28 years old developer.
  19. Ionuț G. Stan · [email protected] · igstan.ro Companion Object class

    Person(name: String, age: Int) { def bio: String = s"$name is a $age years old developer." } object Person { def apply(name: String, age: Int) = { new Person(name, age) } } val person = Person("Ionuț", 28)
  20. Ionuț G. Stan · [email protected] · igstan.ro Companion Object class

    Person(val name: String, val age: Int) { def bio: String = s"$name is a $age years old developer." } object Person { def apply(name: String, age: Int) = { new Person(name, age) } } val person = Person("Ionuț", 28)
  21. Ionuț G. Stan · [email protected] · igstan.ro Case Classes case

    class Person(name: String, age: Int) val p1 = Person("Ionuț", 28) val p2 = Person("Ionuț", 28)
  22. Ionuț G. Stan · [email protected] · igstan.ro Case Classes scala>

    p1.name res1: String = Ionuț scala> p2.age res2: Int = 28 scala> p1 == p2 res3: Boolean = true scala> p1.hashCode == p2.hashCode res4: Boolean = true
  23. Ionuț G. Stan · [email protected] · igstan.ro Pattern Matching val

    s = "bar" s match { case "foo" => 1 case "bar" => 2 case _ => 3 }
  24. Ionuț G. Stan · [email protected] · igstan.ro Pattern Matching case

    class Person(name: String, age: Int) val person = Person("Ionuț", 28) person match { case Person(name, age) => s"$name is $age old." }
  25. Ionuț G. Stan · [email protected] · igstan.ro Pattern Matching case

    class City(name: String, country: String) case class Address(street: String, number: String, city: City) case class Person(name: String, age: Int, address: Address) val p = Person( name = "Ionuț", age = 28, address = Address( number = "42", street = "Endless", city = City("Bucharest", "Romania") ) ) p match { case Person(name, _, Address(_, _, City(city, _))) => s"$name lives in $city." }
  26. Ionuț G. Stan · [email protected] · igstan.ro Pattern Matching case

    class Person(name: String, age: Int) val person = Person("Ionuț", 28) person match { case person: Person => s"${person.name} is ${person.age} old." }
  27. Ionuț G. Stan · [email protected] · igstan.ro A Better Null

    sealed trait Option[+A] case object None extends Option[Nothing] case class Some[A](value: A) extends Option[A] object Option { def apply[A](value: A): Option[A] = { if (value == null) None else Some(value) } }
  28. Ionuț G. Stan · [email protected] · igstan.ro A Better Null

    val user: Option[User] = Users.find(id) user match { case Some(user) => OK(render(user)) case None => NotFound() } // or using higher-order functions user.map(u => OK(render(u))).getOrElse { NotFound() }
  29. Ionuț G. Stan · [email protected] · igstan.ro Pattern Matching "[email protected]"

    match { case Email(user, domain) => println(domain) case _ => println("Invalid email.") }
  30. Ionuț G. Stan · [email protected] · igstan.ro Pattern Matching object

    Email { def unapply(email: String) = { if (valid(email)) { // Another form of pattern matching, in assignment. val Array(user, domain) = email.split('@') Some( (user, domain) ) } else None } private def valid(email: String) = { email.count(char => char == '@') == 1 } } "[email protected]" match { case Email(user, domain) => println(domain) case _ => println("Invalid email.") }
  31. Ionuț G. Stan · [email protected] · igstan.ro OOP vs. FP

    data behaviour behaviour behaviour behaviour behaviour data data data data data
  32. Ionuț G. Stan · [email protected] · igstan.ro OOP behaviour data

    data data data data ‣ one interface defining behaviour ‣ many representations (concrete classes)
  33. Ionuț G. Stan · [email protected] · igstan.ro FP data behaviour

    behaviour behaviour behaviour ‣ one underlying representation (data type) ‣ many behavioural units (functions)
  34. Ionuț G. Stan · [email protected] · igstan.ro OOP vs. FP

    data behaviour behaviour behaviour behaviour behaviour data data data data data
  35. Ionuț G. Stan · [email protected] · igstan.ro Traits trait Comparable[T]

    { def compareTo(other: T): Int def lessThan(other: T): Boolean def equivalent(other: T): Boolean def greaterThan(other: T): Boolean }
  36. Ionuț G. Stan · [email protected] · igstan.ro Traits trait Comparable[T]

    { def compareTo(other: T): Int def lessThan(other: T): Boolean = compareTo(other) < 0 def equivalent(other: T): Boolean = compareTo(other) == 0 def greaterThan(other: T): Boolean = compareTo(other) > 0 }
  37. Ionuț G. Stan · [email protected] · igstan.ro Traits case class

    Seconds(value: Int) extends Comparable[Seconds] { def compareTo(other: Seconds) = value - other.value } Seconds(1).lessThan(Seconds(2)) // true Seconds(1).lessThan(Seconds(1)) // false Seconds(1).greaterThan(Seconds(-2)) // true
  38. Ionuț G. Stan · [email protected] · igstan.ro Traits trait Foo

    { def foo: String = "foo" } trait Bar { def bar: Int = 42 } class Baz extends Foo with Bar val baz = new Baz print(baz.foo) // "foo" print(baz.bar) // 42
  39. Ionuț G. Stan · [email protected] · igstan.ro Traits trait Foo

    { def foo: String = "foo" } trait Bar { def bar: Int = 42 } class Baz val baz = new Baz with Foo with Bar print(baz.foo) // "foo" print(baz.bar) // 42
  40. Ionuț G. Stan · [email protected] · igstan.ro Lazy Evaluation class

    Laziness { lazy val property = { println("Property has been initialized.") 42 } } val a = new Laziness a.property
  41. Ionuț G. Stan · [email protected] · igstan.ro Strict Evaluation def

    first[A,B](a: A, b: B): A = a first(1, 1 / 0)
  42. Ionuț G. Stan · [email protected] · igstan.ro Lazy Evaluation def

    first[A,B](a: A, b: => B): A = a first(1, 1 / 0)
  43. Ionuț G. Stan · [email protected] · igstan.ro Lazy Evaluation def

    double(b: => Int) = b + b double({ println("Evaluating b..."); 2 })
  44. Ionuț G. Stan · [email protected] · igstan.ro Lazy Evaluation def

    unless(cond: Boolean)(body: => Unit) = { if (!cond) body } unless (1 == 2) { println("1 != 2") }
  45. Ionuț G. Stan · [email protected] · igstan.ro Logo Origin Stairs

    from EPFL (École Polytechnique Fédérale de Lausanne), where Scala was born.