Welcome to Swift

Welcome to Swift

1
 Section 1 Swift is a fantastic way to write iOS, macOS, watchOS, and tvOS apps, and will
continue to evolve with new features and capabilities. Our goals for Swift are

About Swift ambitious. We can’t wait to see what you create with it.

About Swift
Swift is a new programming language for iOS, macOS, watchOS, and tvOS apps that
builds on the best of C and Objective-C, without the constraints of C compatibility.
Swift adopts safe programming patterns and adds modern features to make
programming easier, more flexible, and more fun. Swift’s clean slate, backed by the
mature and much-loved Cocoa and Cocoa Touch frameworks, is an opportunity to
reimagine how software development works.

Swift has been years in the making. Apple laid the foundation for Swift by advancing
our existing compiler, debugger, and framework infrastructure. We simplified memory
management with Automatic Reference Counting (ARC). Our framework stack, built
on the solid base of Foundation and Cocoa, has been modernized and standardized
throughout. Objective-C itself has evolved to support blocks, collection literals, and
modules, enabling framework adoption of modern language technologies without
disruption. Thanks to this groundwork, we can now introduce a new language for the
future of Apple software development.

Swift feels familiar to Objective-C developers. It adopts the readability of Objective-C’s

named parameters and the power of Objective-C’s dynamic object model. It provides
seamless access to existing Cocoa frameworks and mix-and-match interoperability
with Objective-C code. Building from this common ground, Swift introduces many new
features and unifies the procedural and object-oriented portions of the language.

Swift is friendly to new programmers. It is the first industrial-quality systems

programming language that is as expressive and enjoyable as a scripting language. It
supports playgrounds, an innovative feature that allows programmers to experiment
with Swift code and see the results immediately, without the overhead of building and
running an app.

Swift combines the best in modern language thinking with wisdom from the wider
Apple engineering culture. The compiler is optimized for performance, and the
language is optimized for development, without compromising on either. It’s designed
to scale from “hello, world” to an entire operating system. All this makes Swift a sound
future investment for developers and for Apple.

2
 Section 2 If the initial value doesn’t provide enough information (or if there is no initial value),
specify the type by writing it after the variable, separated by a colon.

A Swift Tour let implicitInteger = 70

let implicitDouble = 70.0
let explicitDouble: Double = 70
EXPERIMENT
A Swift Tour
Create a constant with an explicit type of Float and a value of 4.
Tradition suggests that the first program in a new language should print the words Values are never implicitly converted to another type. If you need to convert a value to
“Hello, world!” on the screen. In Swift, this can be done in a single line: a different type, explicitly make an instance of the desired type.

print("Hello, world!") let label = "The width is "

If you have written code in C or Objective-C, this syntax looks familiar to you—in let width = 94
Swift, this line of code is a complete program. You don’t need to import a separate let widthLabel = label + String(width)
library for functionality like input/output or string handling. Code written at global EXPERIMENT
scope is used as the entry point for the program, so you don’t need a main() function.
You also don’t need to write semicolons at the end of every statement. Try removing the conversion to String from the last line. What error do you get?

There’s an even simpler way to include values in strings: Write the value in
This tour gives you enough information to start writing code in Swift by showing you parentheses, and write a backslash (\) before the parentheses. For example:
how to accomplish a variety of programming tasks. Don’t worry if you don’t
understand something—everything introduced in this tour is explained in detail in the let apples = 3
rest of this book. let oranges = 5

NOTE let appleSummary = "I have \(apples) apples."

let fruitSummary = "I have \(apples + oranges) pieces of fruit."
On a Mac, download the Playground and double-click the file to open it in Xcode:
https://developer.apple.com/go/?id=swift-tour EXPERIMENT

Use \() to include a floating-point calculation in a string and to include someone’s

Simple Values name in a greeting.
Use let to make a constant and var to make a variable. The value of a constant Create arrays and dictionaries using brackets ([]), and access their elements by
doesn’t need to be known at compile time, but you must assign it a value exactly writing the index or key in brackets. A comma is allowed after the last element.
once. This means you can use constants to name a value that you determine once
but use in many places. var shoppingList = ["catfish", "water", "tulips", "blue paint"]
shoppingList[1] = "bottle of water"
var myVariable = 42
myVariable = 50
var occupations = [
let myConstant = 42
"Malcolm": "Captain",
A constant or variable must have the same type as the value you want to assign to it.
"Kaylee": "Mechanic",
However, you don’t always have to write the type explicitly. Providing a value when
you create a constant or variable lets the compiler infer its type. In the example ]
above, the compiler infers that myVariable is an integer because its initial value is an occupations["Jayne"] = "Public Relations"
integer. To create an empty array or dictionary, use the initializer syntax.

3
 let emptyArray = [String]() Change optionalName to nil. What greeting do you get? Add an else clause that sets
let emptyDictionary = [String: Float]() a different greeting if optionalName is nil.

If type information can be inferred, you can write an empty array as [] and an empty If the optional value is nil, the conditional is false and the code in braces is skipped.
dictionary as [:]—for example, when you set a new value for a variable or pass an Otherwise, the optional value is unwrapped and assigned to the constant after let,
argument to a function. which makes the unwrapped value available inside the block of code.

shoppingList = [] Another way to handle optional values is to provide a default value using the ??
occupations = [:] operator. If the optional value is missing, the default value is used instead.

let nickName: String? = nil

Control Flow
let fullName: String = "John Appleseed"
Use if and switch to make conditionals, and use for-in, for, while, and repeat-
while to make loops. Parentheses around the condition or loop variable are optional. let informalGreeting = "Hi \(nickName ?? fullName)"
Braces around the body are required. Switches support any kind of data and a wide variety of comparison operations—they
aren’t limited to integers and tests for equality.
let individualScores = [75, 43, 103, 87, 12]
var teamScore = 0 let vegetable = "red pepper"
for score in individualScores { switch vegetable {
if score > 50 { case "celery":
teamScore += 3 print("Add some raisins and make ants on a log.")
} else { case "cucumber", "watercress":
teamScore += 1 print("That would make a good tea sandwich.")
} case let x where x.hasSuffix("pepper"):
} print("Is it a spicy \(x)?")
print(teamScore) default:

In an if statement, the conditional must be a Boolean expression—this means that print("Everything tastes good in soup.")
code such as if score { ... } is an error, not an implicit comparison to zero. }
EXPERIMENT
You can use if and let together to work with values that might be missing. These
Try removing the default case. What error do you get?
values are represented as optionals. An optional value either contains a value or
contains nil to indicate that a value is missing. Write a question mark (?) after the Notice how let can be used in a pattern to assign the value that matched the pattern
type of a value to mark the value as optional. to a constant.

var optionalString: String? = "Hello" After executing the code inside the switch case that matched, the program exits from
print(optionalString == nil) the switch statement. Execution doesn’t continue to the next case, so there is no need
to explicitly break out of the switch at the end of each case’s code.
var optionalName: String? = "John Appleseed"
You use for-in to iterate over items in a dictionary by providing a pair of names to use
var greeting = "Hello!"
for each key-value pair. Dictionaries are an unordered collection, so their keys and
if let name = optionalName { values are iterated over in an arbitrary order.
greeting = "Hello, \(name)"
} let interestingNumbers = [

EXPERIMENT
"Prime": [2, 3, 5, 7, 11, 13],

4
 "Fibonacci": [1, 1, 2, 3, 5, 8],
Functions and Closures
"Square": [1, 4, 9, 16, 25],
Use func to declare a function. Call a function by following its name with a list of
] arguments in parentheses. Use -> to separate the parameter names and types from
var largest = 0 the function’s return type.
for (kind, numbers) in interestingNumbers {
func greet(person: String, day: String) -> String {
for number in numbers {
return "Hello \(person), today is \(day)."
if number > largest {
}
largest = number
greet(person: "Bob", day: "Tuesday")
}
EXPERIMENT
}
Remove the day parameter. Add a parameter to include today’s lunch special in the
}
greeting.
print(largest)
EXPERIMENT
By default, functions use their parameter names as labels for their arguments. Write a
custom argument label before the parameter name, or write _ to use no argument
Add another variable to keep track of which kind of number was the largest, as well as label.
what that largest number was.

Use while to repeat a block of code until a condition changes. The condition of a loop func greet(_ person: String, on day: String) -> String {
can be at the end instead, ensuring that the loop is run at least once. return "Hello \(person), today is \(day)."
}
var n = 2
greet("John", on: "Wednesday")
while n < 100 {
Use a tuple to make a compound value—for example, to return multiple values from a
n = n * 2
function. The elements of a tuple can be referred to either by name or by number.
}
print(n) func calculateStatistics(scores: [Int]) -> (min: Int, max: Int,
sum: Int) {
var min = scores[0]
var m = 2
var max = scores[0]
repeat {
var sum = 0
m = m * 2
} while m < 100
for score in scores {
print(m)
if score > max {
You can keep an index in a loop by using ..< to make a range of indexes.
max = score

var total = 0 } else if score < min {

for i in 0..<4 { min = score

total += i }

} sum += score

print(total) }

Use ..< to make a range that omits its upper value, and use ... to make a range that
includes both values. return (min, max, sum)
}

5
 let statistics = calculateStatistics(scores: [5, 3, 100, 3, 9]) increment(7)
print(statistics.sum) A function can take another function as one of its arguments.
print(statistics.2)
func hasAnyMatches(list: [Int], condition: (Int) -> Bool) -> Bool {
Functions can also take a variable number of arguments, collecting them into an array.
for item in list {
func sumOf(numbers: Int...) -> Int { if condition(item) {
var sum = 0 return true
for number in numbers { }
sum += number }
} return false
return sum }
} func lessThanTen(number: Int) -> Bool {
sumOf() return number < 10
sumOf(numbers: 42, 597, 12) }
EXPERIMENT var numbers = [20, 19, 7, 12]

Write a function that calculates the average of its arguments. hasAnyMatches(list: numbers, condition: lessThanTen)

Functions can be nested. Nested functions have access to variables that were Functions are actually a special case of closures: blocks of code that can be called
declared in the outer function. You can use nested functions to organize the code in a later. The code in a closure has access to things like variables and functions that were
function that is long or complex. available in the scope where the closure was created, even if the closure is in a
different scope when it is executed—you saw an example of this already with nested
func returnFifteen() -> Int { functions. You can write a closure without a name by surrounding code with braces
({}). Use in to separate the arguments and return type from the body.
var y = 10
func add() { numbers.map({
y += 5 (number: Int) -> Int in
} let result = 3 * number
add() return result
return y })
} EXPERIMENT
returnFifteen()
Rewrite the closure to return zero for all odd numbers.
Functions are a first-class type. This means that a function can return another function
You have several options for writing closures more concisely. When a closure’s type is
as its value.
already known, such as the callback for a delegate, you can omit the type of its
func makeIncrementer() -> ((Int) -> Int) { parameters, its return type, or both. Single statement closures implicitly return the
value of their only statement.
func addOne(number: Int) -> Int {
return 1 + number let mappedNumbers = numbers.map({ number in 3 * number })
} print(mappedNumbers)
return addOne You can refer to parameters by number instead of by name—this approach is
} especially useful in very short closures. A closure passed as the last argument to a
var increment = makeIncrementer()

6
function can appear immediately after the parentheses. When a closure is the only }
argument to a function, you can omit the parentheses entirely. Notice how self is used to distinguish the name property from the name argument to
the initializer. The arguments to the initializer are passed like a function call when you
let sortedNumbers = numbers.sorted { $0 > $1 } create an instance of the class. Every property needs a value assigned—either in its
print(sortedNumbers) declaration (as with numberOfSides) or in the initializer (as with name).

Objects and Classes Use deinit to create a deinitializer if you need to perform some cleanup before the
Use class followed by the class’s name to create a class. A property declaration in a object is deallocated.
class is written the same way as a constant or variable declaration, except that it is in
the context of a class. Likewise, method and function declarations are written the Subclasses include their superclass name after their class name, separated by a
same way. colon. There is no requirement for classes to subclass any standard root class, so you
can include or omit a superclass as needed.
class Shape {
var numberOfSides = 0 Methods on a subclass that override the superclass’s implementation are marked with
override—overriding a method by accident, without override, is detected by the
func simpleDescription() -> String {
compiler as an error. The compiler also detects methods with override that don’t
return "A shape with \(numberOfSides) sides."
actually override any method in the superclass.
}
} class Square: NamedShape {

EXPERIMENT var sideLength: Double

Add a constant property with let, and add another method that takes an argument.
init(sideLength: Double, name: String) {
Create an instance of a class by putting parentheses after the class name. Use dot
self.sideLength = sideLength
syntax to access the properties and methods of the instance.
super.init(name: name)
var shape = Shape() numberOfSides = 4
shape.numberOfSides = 7 }
var shapeDescription = shape.simpleDescription()

This version of the Shape class is missing something important: an initializer to set up func area() -> Double {
the class when an instance is created. Use init to create one. return sideLength * sideLength
}
class NamedShape {
var numberOfSides: Int = 0
override func simpleDescription() -> String {
var name: String
return "A square with sides of length \(sideLength)."
}
init(name: String) {
}
self.name = name
let test = Square(sideLength: 5.2, name: "my test square")
}
test.area()
test.simpleDescription()
func simpleDescription() -> String {
EXPERIMENT
return "A shape with \(numberOfSides) sides."
}

7
 Make another subclass of NamedShape called Circle that takes a radius and a name 2.Calling the superclass’s initializer.
as arguments to its initializer. Implement an area() and a simpleDescription()
method on the Circle class. 3.Changing the value of properties defined by the superclass. Any additional
setup work that uses methods, getters, or setters can also be done at this
In addition to simple properties that are stored, properties can have a getter and a point.
setter.
If you don’t need to compute the property but still need to provide code that is run
class EquilateralTriangle: NamedShape { before and after setting a new value, use willSet and didSet. The code you provide
var sideLength: Double = 0.0
is run any time the value changes outside of an initializer. For example, the class
below ensures that the side length of its triangle is always the same as the side length
of its square.
init(sideLength: Double, name: String) {
self.sideLength = sideLength class TriangleAndSquare {
super.init(name: name) var triangle: EquilateralTriangle {
numberOfSides = 3 willSet {
} square.sideLength = newValue.sideLength
}
var perimeter: Double { }
get { var square: Square {
return 3.0 * sideLength willSet {
} triangle.sideLength = newValue.sideLength
set { }
sideLength = newValue / 3.0 }
} init(size: Double, name: String) {
} square = Square(sideLength: size, name: name)
triangle = EquilateralTriangle(sideLength: size, name: name)
override func simpleDescription() -> String { }
return "An equilateral triangle with sides of length \ }
(sideLength)." var triangleAndSquare = TriangleAndSquare(size: 10, name: "another
} test shape")
} print(triangleAndSquare.square.sideLength)
var triangle = EquilateralTriangle(sideLength: 3.1, name: "a print(triangleAndSquare.triangle.sideLength)
triangle") triangleAndSquare.square = Square(sideLength: 50, name: "larger
print(triangle.perimeter) square")
triangle.perimeter = 9.9 print(triangleAndSquare.triangle.sideLength)
print(triangle.sideLength) When working with optional values, you can write ? before operations like methods,
In the setter for perimeter, the new value has the implicit name newValue. You can properties, and subscripting. If the value before the ? is nil, everything after the ? is
provide an explicit name in parentheses after set. ignored and the value of the whole expression is nil. Otherwise, the optional value is
unwrapped, and everything after the ? acts on the unwrapped value. In both cases,
the value of the whole expression is an optional value.
Notice that the initializer for the EquilateralTriangle class has three different steps:
let optionalSquare: Square? = Square(sideLength: 2.5, name:
1.Setting the value of properties that the subclass declares. "optional square")

8
 let sideLength = optionalSquare?.sideLength let threeDescription = convertedRank.simpleDescription()
}
Enumerations and Structures The case values of an enumeration are actual values, not just another way of writing
Use enum to create an enumeration. Like classes and all other named types, their raw values. In fact, in cases where there isn’t a meaningful raw value, you don’t
enumerations can have methods associated with them. have to provide one.

enum Rank: Int { enum Suit {

case ace = 1 case spades, hearts, diamonds, clubs
case two, three, four, five, six, seven, eight, nine, ten func simpleDescription() -> String {
case jack, queen, king switch self {
func simpleDescription() -> String { case .spades:
switch self { return "spades"
case .ace: case .hearts:
return "ace" return "hearts"
case .jack: case .diamonds:
return "jack" return "diamonds"
case .queen: case .clubs:
return "queen" return "clubs"
case .king: }
return "king" }
default: }
return String(self.rawValue) let hearts = Suit.hearts
} let heartsDescription = hearts.simpleDescription()
} EXPERIMENT
}
Add a color() method to Suit that returns “black” for spades and clubs, and returns
let ace = Rank.ace “red” for hearts and diamonds.
let aceRawValue = ace.rawValue
Notice the two ways that the hearts case of the enumeration is referred to above:
EXPERIMENT When assigning a value to the hearts constant, the enumeration case Suit.hearts is
Write a function that compares two Rank values by comparing their raw values. referred to by its full name because the constant doesn’t have an explicit type
specified. Inside the switch, the enumeration case is referred to by the abbreviated
By default, Swift assigns the raw values starting at zero and incrementing by one each form .hearts because the value of self is already known to be a suit. You can use
time, but you can change this behavior by explicitly specifying values. In the example the abbreviated form anytime the value’s type is already known.
above, Ace is explicitly given a raw value of 1, and the rest of the raw values are
assigned in order. You can also use strings or floating-point numbers as the raw type If an enumeration has raw values, those values are determined as part of the
of an enumeration. Use the rawValue property to access the raw value of an declaration, which means every instance of a particular enumeration case always has
enumeration case. the same raw value. Another choice for enumeration cases is to have values
associated with the case—these values are determined when you make the instance,
Use the init?(rawValue:) initializer to make an instance of an enumeration from a and they can be different for each instance of an enumeration case. You can think of
raw value. the associated values as behaving like stored properties of the enumeration case
instance. For example, consider the case of requesting the sunrise and sunset times
if let convertedRank = Rank(rawValue: 3) {

9
from a server. The server either responds with the requested information, or it Protocols and Extensions
responds with a description of what went wrong.
Use protocol to declare a protocol.
enum ServerResponse {
protocol ExampleProtocol {
case result(String, String)
var simpleDescription: String { get }
case failure(String)
mutating func adjust()
}
}

Classes, enumerations, and structs can all adopt protocols.

let success = ServerResponse.result("6:00 am", "8:09 pm")
let failure = ServerResponse.failure("Out of cheese.") class SimpleClass: ExampleProtocol {
var simpleDescription: String = "A very simple class."
switch success { var anotherProperty: Int = 69105
case let .result(sunrise, sunset): func adjust() {
print("Sunrise is at \(sunrise) and sunset is at \(sunset).") simpleDescription += " Now 100% adjusted."
case let .failure(message): }
print("Failure... \(message)") }
} var a = SimpleClass()
EXPERIMENT a.adjust()
Add a third case to ServerResponse and to the switch. let aDescription = a.simpleDescription

Notice how the sunrise and sunset times are extracted from the ServerResponse value
as part of matching the value against the switch cases. struct SimpleStructure: ExampleProtocol {
var simpleDescription: String = "A simple structure"
Use struct to create a structure. Structures support many of the same behaviors as mutating func adjust() {
classes, including methods and initializers. One of the most important differences simpleDescription += " (adjusted)"
between structures and classes is that structures are always copied when they are
}
passed around in your code, but classes are passed by reference.
}
struct Card { var b = SimpleStructure()
var rank: Rank b.adjust()
var suit: Suit let bDescription = b.simpleDescription
func simpleDescription() -> String { EXPERIMENT
return "The \(rank.simpleDescription()) of \ Write an enumeration that conforms to this protocol.
(suit.simpleDescription())"
} Notice the use of the mutating keyword in the declaration of SimpleStructure to mark
}
a method that modifies the structure. The declaration of SimpleClass doesn’t need
any of its methods marked as mutating because methods on a class can always
let threeOfSpades = Card(rank: .three, suit: .spades)
modify the class.
let threeOfSpadesDescription = threeOfSpades.simpleDescription()
EXPERIMENT Use extension to add functionality to an existing type, such as new methods and
Add a method to Card that creates a full deck of cards, with one card of each
computed properties. You can use an extension to add protocol conformance to a type
combination of rank and suit.

10
that is declared elsewhere, or even to a type that you imported from a library or func send(job: Int, toPrinter printerName: String) throws -> String
{
framework.
if printerName == "Never Has Toner" {
extension Int: ExampleProtocol { throw PrinterError.noToner
var simpleDescription: String { }
return "The number \(self)" return "Job sent"
} }
mutating func adjust() { There are several ways to handle errors. One way is to use do-catch. Inside the do
self += 42 block, you mark code that can throw an error by writing try in front of it. Inside the
} catch block, the error is automatically given the name error unless you give it a
} different name.
print(7.simpleDescription)
do {
EXPERIMENT
let printerResponse = try send(job: 1040, toPrinter: "Bi
Write an extension for the Double type that adds an absoluteValue property. Sheng")
print(printerResponse)
You can use a protocol name just like any other named type—for example, to create a
} catch {
collection of objects that have different types but that all conform to a single protocol.
When you work with values whose type is a protocol type, methods outside the print(error)
protocol definition are not available. }
EXPERIMENT
let protocolValue: ExampleProtocol = a
Change the printer name to "Never Has Toner", so that the send(job:toPrinter:)
print(protocolValue.simpleDescription)
function throws an error.
// print(protocolValue.anotherProperty) // Uncomment to see the
error You can provide multiple catch blocks that handle specific errors. You write a pattern
after catch just as you do after case in a switch.
Even though the variable protocolValue has a runtime type of SimpleClass, the
compiler treats it as the given type of ExampleProtocol. This means that you can’t
do {
accidentally access methods or properties that the class implements in addition to its
protocol conformance. let printerResponse = try send(job: 1440, toPrinter:
"Gutenberg")
print(printerResponse)
Error Handling
} catch PrinterError.onFire {
You represent errors using any type that adopts the Error protocol.
print("I'll just put this over here, with the rest of the
fire.")
enum PrinterError: Error {
} catch let printerError as PrinterError {
case outOfPaper
print("Printer error: \(printerError).")
case noToner
} catch {
case onFire
print(error)
}
}
Use throw to throw an error and throws to mark a function that can throw an error. If
EXPERIMENT
you throw an error in a function, the function returns immediately and the code that
called the function handles the error. Add code to throw an error inside the do block. What kind of error do you need to throw
so that the error is handled by the first catch block? What about the second and third
blocks?

11
Another way to handle errors is to use try? to convert the result to an optional. If the You can make generic forms of functions and methods, as well as classes,
function throws an error, the specific error is discarded and the result is nil. enumerations, and structures.
Otherwise, the result is an optional containing the value that the function returned.
// Reimplement the Swift standard library's optional type
let printerSuccess = try? send(job: 1884, toPrinter: enum OptionalValue<Wrapped> {
"Mergenthaler")
case none
let printerFailure = try? send(job: 1885, toPrinter: "Never Has
Toner") case some(Wrapped)

Use defer to write a block of code that is executed after all other code in the function, }
just before the function returns. The code is executed regardless of whether the var possibleInteger: OptionalValue<Int> = .none
function throws an error. You can use defer to write setup and cleanup code next to possibleInteger = .some(100)
each other, even though they need to be executed at different times.
Use where right before the body to specify a list of requirements—for example, to
var fridgeIsOpen = false
require the type to implement a protocol, to require two types to be the same, or to
require a class to have a particular superclass.
let fridgeContent = ["milk", "eggs", "leftovers"]
func anyCommonElements<T: Sequence, U: Sequence>(_ lhs: T, _ rhs:
U) -> Bool
func fridgeContains(_ food: String) -> Bool {
where T.Iterator.Element: Equatable, T.Iterator.Element ==
fridgeIsOpen = true
U.Iterator.Element {
defer {
for lhsItem in lhs {
fridgeIsOpen = false
for rhsItem in rhs {
}
if lhsItem == rhsItem {
return true
let result = fridgeContent.contains(food)
}
return result
}
}
}
fridgeContains("banana")
return false
print(fridgeIsOpen)
}
anyCommonElements([1, 2, 3], [3])
Generics
EXPERIMENT
Write a name inside angle brackets to make a generic function or type.
Modify the anyCommonElements(_:_:) function to make a function that returns an
func makeArray<Item>(repeating item: Item, numberOfTimes: Int) -> array of the elements that any two sequences have in common.
[Item] {
Writing <T: Equatable> is the same as writing <T> ... where T: Equatable>.
var result = [Item]()
for _ in 0..<numberOfTimes {
result.append(item)
}
return result
}
makeArray(repeating: "knock", numberOfTimes:4)

12
 Language Guide
Language Guide

13
 Section 1 Constants and variables associate a name (such as maximumNumberOfLoginAttempts
or welcomeMessage) with a value of a particular type (such as the number 10 or the

The Basics string "Hello"). The value of a constant cannot be changed once it is set, whereas a
variable can be set to a different value in the future.

Declaring Constants and Variables

Constants and variables must be declared before they are used. You declare
The Basics constants with the let keyword and variables with the var keyword. Here’s an
example of how constants and variables can be used to track the number of login
Swift is a new programming language for iOS, macOS, watchOS, and tvOS app attempts a user has made:
development. Nonetheless, many parts of Swift will be familiar from your experience
of developing in C and Objective-C. let maximumNumberOfLoginAttempts = 10
var currentLoginAttempt = 0
Swift provides its own versions of all fundamental C and Objective-C types, including This code can be read as:
Int for integers, Double and Float for floating-point values, Bool for Boolean values,
and String for textual data. Swift also provides powerful versions of the three primary
“Declare a new constant called maximumNumberOfLoginAttempts, and give it a value of
collection types, Array, Set, and Dictionary, as described in Collection Types.
10. Then, declare a new variable called currentLoginAttempt, and give it an initial
value of 0.”
Like C, Swift uses variables to store and refer to values by an identifying name. Swift
also makes extensive use of variables whose values cannot be changed. These are
In this example, the maximum number of allowed login attempts is declared as a
known as constants, and are much more powerful than constants in C. Constants are
constant, because the maximum value never changes. The current login attempt
used throughout Swift to make code safer and clearer in intent when you work with
counter is declared as a variable, because this value must be incremented after each
values that do not need to change.
failed login attempt.

In addition to familiar types, Swift introduces advanced types not found in Objective-C,
You can declare multiple constants or multiple variables on a single line, separated by
such as tuples. Tuples enable you to create and pass around groupings of values.
commas:
You can use a tuple to return multiple values from a function as a single compound
value. var x = 0.0, y = 0.0, z = 0.0
NOTE
Swift also introduces optional types, which handle the absence of a value. Optionals
say either “there is a value, and it equals x” or “there isn’t a value at all”. Using If a stored value in your code is not going to change, always declare it as a constant
optionals is similar to using nil with pointers in Objective-C, but they work for any with the let keyword. Use variables only for storing values that need to be able to
change.
type, not just classes. Not only are optionals safer and more expressive than nil
pointers in Objective-C, they are at the heart of many of Swift’s most powerful
Type Annotations
features.
You can provide a type annotation when you declare a constant or variable, to be
clear about the kind of values the constant or variable can store. Write a type
Swift is a type-safe language, which means the language helps you to be clear about
annotation by placing a colon after the constant or variable name, followed by a
the types of values your code can work with. If part of your code expects a String,
space, followed by the name of the type to use.
type safety prevents you from passing it an Int by mistake. Likewise, type safety
prevents you from accidentally passing an optional String to a piece of code that
expects a nonoptional String. Type safety helps you catch and fix errors as early as This example provides a type annotation for a variable called welcomeMessage, to
possible in the development process. indicate that the variable can store String values:

var welcomeMessage: String

Constants and Variables
The colon in the declaration means “…of type…,” so the code above can be read as:

14
“Declare a variable called welcomeMessage that is of type String.” var friendlyWelcome = "Hello!"
friendlyWelcome = "Bonjour!"
The phrase “of type String” means “can store any String value.” Think of it as // friendlyWelcome is now "Bonjour!"
meaning “the type of thing” (or “the kind of thing”) that can be stored.
Unlike a variable, the value of a constant cannot be changed once it is set. Attempting
to do so is reported as an error when your code is compiled:
The welcomeMessage variable can now be set to any string value without error:
let languageName = "Swift"
welcomeMessage = "Hello"
languageName = "Swift++"
You can define multiple related variables of the same type on a single line, separated
// This is a compile-time error: languageName cannot be changed.
by commas, with a single type annotation after the final variable name:
Printing Constants and Variables
var red, green, blue: Double
You can print the current value of a constant or variable with the
NOTE
print(_:separator:terminator:) function:
It is rare that you need to write type annotations in practice. If you provide an initial
value for a constant or variable at the point that it is defined, Swift can almost always print(friendlyWelcome)
infer the type to be used for that constant or variable, as described in Type Safety and
// Prints "Bonjour!"
Type Inference. In the welcomeMessage example above, no initial value is provided,
and so the type of the welcomeMessage variable is specified with a type annotation The print(_:separator:terminator:) function is a global function that prints one or
rather than being inferred from an initial value. more values to an appropriate output. In Xcode, for example, the
print(_:separator:terminator:) function prints its output in Xcode’s “console”
Naming Constants and Variables
pane. The separator and terminator parameter have default values, so you can omit
Constant and variable names can contain almost any character, including Unicode them when you call this function. By default, the function terminates the line it prints by
characters: adding a line break. To print a value without a line break after it, pass an empty string
as the terminator—for example, print(someValue, terminator: ""). For information
let π = 3.14159 about parameters with default values, see Default Parameter Values.
let 你好 = "你好世界"
Swift uses string interpolation to include the name of a constant or variable as a
let 🐶 🐮 = "dogcow"
placeholder in a longer string, and to prompt Swift to replace it with the current value
Constant and variable names cannot contain whitespace characters, mathematical of that constant or variable. Wrap the name in parentheses and escape it with a
symbols, arrows, private-use (or invalid) Unicode code points, or line- and box- backslash before the opening parenthesis:
drawing characters. Nor can they begin with a number, although numbers may be
included elsewhere within the name. print("The current value of friendlyWelcome is \(friendlyWelcome)")
// Prints "The current value of friendlyWelcome is Bonjour!"
Once you’ve declared a constant or variable of a certain type, you can’t redeclare it NOTE
again with the same name, or change it to store values of a different type. Nor can you
change a constant into a variable or a variable into a constant. All options you can use with string interpolation are described in String Interpolation.

NOTE Comments
If you need to give a constant or variable the same name as a reserved Swift keyword, Use comments to include nonexecutable text in your code, as a note or reminder to
surround the keyword with backticks (`) when using it as a name. However, avoid using yourself. Comments are ignored by the Swift compiler when your code is compiled.
keywords as names unless you have absolutely no choice.

You can change the value of an existing variable to another value of a compatible Comments in Swift are very similar to comments in C. Single-line comments begin
type. In this example, the value of friendlyWelcome is changed from "Hello!" to with two forward-slashes (//):
"Bonjour!":
// This is a comment.

15
Multiline comments start with a forward-slash followed by an asterisk (/*) and end The values of these properties are of the appropriate-sized number type (such as
with an asterisk followed by a forward-slash (*/): UInt8 in the example above) and can therefore be used in expressions alongside
other values of the same type.
/* This is also a comment
but is written over multiple lines. */ Int
Unlike multiline comments in C, multiline comments in Swift can be nested inside In most cases, you don’t need to pick a specific size of integer to use in your code.
other multiline comments. You write nested comments by starting a multiline comment Swift provides an additional integer type, Int, which has the same size as the current
block and then starting a second multiline comment within the first block. The second platform’s native word size:
block is then closed, followed by the first block:
On a 32-bit platform, Int is the same size as Int32.
/* This is the start of the first multiline comment.
On a 64-bit platform, Int is the same size as Int64.
/* This is the second, nested multiline comment. */
Unless you need to work with a specific size of integer, always use Int for integer
This is the end of the first multiline comment. */
values in your code. This aids code consistency and interoperability. Even on 32-bit
Nested multiline comments enable you to comment out large blocks of code quickly platforms, Int can store any value between -2,147,483,648 and 2,147,483,647, and
and easily, even if the code already contains multiline comments. is large enough for many integer ranges.

Semicolons UInt
Unlike many other languages, Swift does not require you to write a semicolon (;) after Swift also provides an unsigned integer type, UInt, which has the same size as the
each statement in your code, although you can do so if you wish. However, current platform’s native word size:
semicolons are required if you want to write multiple separate statements on a single
line: On a 32-bit platform, UInt is the same size as UInt32.
On a 64-bit platform, UInt is the same size as UInt64.
let cat = "🐱 "; print(cat)
NOTE
// Prints "🐱 " Use UInt only when you specifically need an unsigned integer type with the same size
as the platform’s native word size. If this is not the case, Int is preferred, even when
Integers the values to be stored are known to be non-negative. A consistent use of Int for
integer values aids code interoperability, avoids the need to convert between different
Integers are whole numbers with no fractional component, such as 42 and -23. number types, and matches integer type inference, as described in Type Safety and
Integers are either signed (positive, zero, or negative) or unsigned (positive or zero). Type Inference.

Swift provides signed and unsigned integers in 8, 16, 32, and 64 bit forms. These Floating-Point Numbers
integers follow a naming convention similar to C, in that an 8-bit unsigned integer is of Floating-point numbers are numbers with a fractional component, such as 3.14159,
type UInt8, and a 32-bit signed integer is of type Int32. Like all types in Swift, these 0.1, and -273.15.
integer types have capitalized names.
Floating-point types can represent a much wider range of values than integer types,
Integer Bounds and can store numbers that are much larger or smaller than can be stored in an Int.
You can access the minimum and maximum values of each integer type with its min Swift provides two signed floating-point number types:
and max properties:
Double represents a 64-bit floating-point number.
let minValue = UInt8.min // minValue is equal to 0, and is of type
UInt8 Float represents a 32-bit floating-point number.
let maxValue = UInt8.max // maxValue is equal to 255, and is of NOTE
type UInt8

16
 Double has a precision of at least 15 decimal digits, whereas the precision of Float If you combine integer and floating-point literals in an expression, a type of Double will
can be as little as 6 decimal digits. The appropriate floating-point type to use depends be inferred from the context:
on the nature and range of values you need to work with in your code. In situations
where either type would be appropriate, Double is preferred.
let anotherPi = 3 + 0.14159
// anotherPi is also inferred to be of type Double
Type Safety and Type Inference
Swift is a type-safe language. A type safe language encourages you to be clear about The literal value of 3 has no explicit type in and of itself, and so an appropriate output
the types of values your code can work with. If part of your code expects a String, type of Double is inferred from the presence of a floating-point literal as part of the
you can’t pass it an Int by mistake. addition.

Because Swift is type safe, it performs type checks when compiling your code and Numeric Literals
flags any mismatched types as errors. This enables you to catch and fix errors as Integer literals can be written as:
early as possible in the development process.
A decimal number, with no prefix
Type-checking helps you avoid errors when you’re working with different types of
A binary number, with a 0b prefix
values. However, this doesn’t mean that you have to specify the type of every
constant and variable that you declare. If you don’t specify the type of value you need, An octal number, with a 0o prefix
Swift uses type inference to work out the appropriate type. Type inference enables a A hexadecimal number, with a 0x prefix
compiler to deduce the type of a particular expression automatically when it compiles
your code, simply by examining the values you provide. All of these integer literals have a decimal value of 17:

let decimalInteger = 17
Because of type inference, Swift requires far fewer type declarations than languages
such as C or Objective-C. Constants and variables are still explicitly typed, but much let binaryInteger = 0b10001 // 17 in binary notation
of the work of specifying their type is done for you. let octalInteger = 0o21 // 17 in octal notation
let hexadecimalInteger = 0x11 // 17 in hexadecimal notation
Type inference is particularly useful when you declare a constant or variable with an
Floating-point literals can be decimal (with no prefix), or hexadecimal (with a 0x
initial value. This is often done by assigning a literal value (or literal) to the constant or
prefix). They must always have a number (or hexadecimal number) on both sides of
variable at the point that you declare it. (A literal value is a value that appears directly
the decimal point. Decimal floats can also have an optional exponent, indicated by an
in your source code, such as 42 and 3.14159 in the examples below.)
uppercase or lowercase e; hexadecimal floats must have an exponent, indicated by an
uppercase or lowercase p.
For example, if you assign a literal value of 42 to a new constant without saying what
type it is, Swift infers that you want the constant to be an Int, because you have
For decimal numbers with an exponent of exp, the base number is multiplied by 10exp:
initialized it with a number that looks like an integer:

let meaningOfLife = 42 1.25e2 means 1.25 x 102, or 125.0.

// meaningOfLife is inferred to be of type Int 1.25e-2 means 1.25 x 10-2, or 0.0125.

Likewise, if you don’t specify a type for a floating-point literal, Swift infers that you For hexadecimal numbers with an exponent of exp, the base number is multiplied by
want to create a Double: 2exp:

let pi = 3.14159 0xFp2 means 15 x 22, or 60.0.

// pi is inferred to be of type Double
0xFp-2 means 15 x 2-2, or 3.75.
Swift always chooses Double (rather than Float) when inferring the type of floating- All of these floating-point literals have a decimal value of 12.1875:
point numbers.
let decimalDouble = 12.1875

17
 let exponentDouble = 1.21875e1 is of type UInt16, whereas the constant one is of type UInt8. They cannot be added
let hexadecimalDouble = 0xC.3p0 together directly, because they are not of the same type. Instead, this example calls
UInt16(one) to create a new UInt16 initialized with the value of one, and uses this
Numeric literals can contain extra formatting to make them easier to read. Both
value in place of the original:
integers and floats can be padded with extra zeros and can contain underscores to
help with readability. Neither type of formatting affects the underlying value of the
let twoThousand: UInt16 = 2_000
literal:
let one: UInt8 = 1
let paddedDouble = 000123.456 let twoThousandAndOne = twoThousand + UInt16(one)
let oneMillion = 1_000_000 Because both sides of the addition are now of type UInt16, the addition is allowed.
let justOverOneMillion = 1_000_000.000_000_1 The output constant (twoThousandAndOne) is inferred to be of type UInt16, because it
is the sum of two UInt16 values.
Numeric Type Conversion
SomeType(ofInitialValue) is the default way to call the initializer of a Swift type and
Use the Int type for all general-purpose integer constants and variables in your code,
even if they are known to be non-negative. Using the default integer type in everyday pass in an initial value. Behind the scenes, UInt16 has an initializer that accepts a
UInt8 value, and so this initializer is used to make a new UInt16 from an existing
situations means that integer constants and variables are immediately interoperable in
UInt8. You can’t pass in any type here, however—it has to be a type for which UInt16
your code and will match the inferred type for integer literal values.
provides an initializer. Extending existing types to provide initializers that accept new
types (including your own type definitions) is covered in Extensions.
Use other integer types only when they are specifically needed for the task at hand,
because of explicitly-sized data from an external source, or for performance, memory
usage, or other necessary optimization. Using explicitly-sized types in these situations
Integer and Floating-Point Conversion
helps to catch any accidental value overflows and implicitly documents the nature of Conversions between integer and floating-point numeric types must be made explicit:
the data being used.
let three = 3
Integer Conversion let pointOneFourOneFiveNine = 0.14159

The range of numbers that can be stored in an integer constant or variable is different let pi = Double(three) + pointOneFourOneFiveNine
for each numeric type. An Int8 constant or variable can store numbers between -128 // pi equals 3.14159, and is inferred to be of type Double
and 127, whereas a UInt8 constant or variable can store numbers between 0 and 255.
Here, the value of the constant three is used to create a new value of type Double, so
A number that will not fit into a constant or variable of a sized integer type is reported
that both sides of the addition are of the same type. Without this conversion in place,
as an error when your code is compiled:
the addition would not be allowed.
let cannotBeNegative: UInt8 = -1
Floating-point to integer conversion must also be made explicit. An integer type can be
// UInt8 cannot store negative numbers, and so this will report an
error
initialized with a Double or Float value:
let tooBig: Int8 = Int8.max + 1 let integerPi = Int(pi)
// Int8 cannot store a number larger than its maximum value, // integerPi equals 3, and is inferred to be of type Int
// and so this will also report an error
Floating-point values are always truncated when used to initialize a new integer value
Because each numeric type can store a different range of values, you must opt in to in this way. This means that 4.75 becomes 4, and -3.9 becomes -3.
numeric type conversion on a case-by-case basis. This opt-in approach prevents
hidden conversion errors and helps make type conversion intentions explicit in your NOTE
code.
The rules for combining numeric constants and variables are different from the rules for
numeric literals. The literal value 3 can be added directly to the literal value 0.14159,
To convert one specific number type to another, you initialize a new number of the because number literals do not have an explicit type in and of themselves. Their type is
desired type with the existing value. In the example below, the constant twoThousand inferred only at the point that they are evaluated by the compiler.

18
Type Aliases Conditional statements such as the if statement are covered in more detail in Control
Flow.
Type aliases define an alternative name for an existing type. You define type aliases
with the typealias keyword.
Swift’s type safety prevents non-Boolean values from being substituted for Bool. The
following example reports a compile-time error:
Type aliases are useful when you want to refer to an existing type by a name that is
contextually more appropriate, such as when working with data of a specific size from
let i = 1
an external source:
if i {
typealias AudioSample = UInt16 // this example will not compile, and will report an error

Once you define a type alias, you can use the alias anywhere you might use the }
original name: However, the alternative example below is valid:

var maxAmplitudeFound = AudioSample.min let i = 1

// maxAmplitudeFound is now 0 if i == 1 {
Here, AudioSample is defined as an alias for UInt16. Because it is an alias, the call to // this example will compile successfully
AudioSample.min actually calls UInt16.min, which provides an initial value of 0 for the }
maxAmplitudeFound variable.
The result of the i == 1 comparison is of type Bool, and so this second example
passes the type-check. Comparisons like i == 1 are discussed in Basic Operators.
Booleans
Swift has a basic Boolean type, called Bool. Boolean values are referred to as logical, As with other examples of type safety in Swift, this approach avoids accidental errors
because they can only ever be true or false. Swift provides two Boolean constant and ensures that the intention of a particular section of code is always clear.
values, true and false:

let orangesAreOrange = true

Tuples
let turnipsAreDelicious = false
Tuples group multiple values into a single compound value. The values within a tuple
can be of any type and do not have to be of the same type as each other.
The types of orangesAreOrange and turnipsAreDelicious have been inferred as Bool
from the fact that they were initialized with Boolean literal values. As with Int and In this example, (404, "Not Found") is a tuple that describes an HTTP status code.
Double above, you don’t need to declare constants or variables as Bool if you set An HTTP status code is a special value returned by a web server whenever you
them to true or false as soon as you create them. Type inference helps make Swift request a web page. A status code of 404 Not Found is returned if you request a
code more concise and readable when it initializes constants or variables with other webpage that doesn’t exist.
values whose type is already known.
let http404Error = (404, "Not Found")
Boolean values are particularly useful when you work with conditional statements // http404Error is of type (Int, String), and equals (404, "Not
such as the if statement: Found")

if turnipsAreDelicious {
The (404, "Not Found") tuple groups together an Int and a String to give the HTTP
status code two separate values: a number and a human-readable description. It can
print("Mmm, tasty turnips!")
be described as “a tuple of type (Int, String)”.
} else {
print("Eww, turnips are horrible.") You can create tuples from any permutation of types, and they can contain as many
} different types as you like. There’s nothing stopping you from having a tuple of type
(Int, Int, Int), or (String, Bool), or indeed any other permutation you require.
// Prints "Eww, turnips are horrible."

19
You can decompose a tuple’s contents into separate constants or variables, which you temporary scope, model it as a class or structure, rather than as a tuple. For more
then access as usual: information, see Classes and Structures.

let (statusCode, statusMessage) = http404Error Optionals

print("The status code is \(statusCode)") You use optionals in situations where a value may be absent. An optional represents
// Prints "The status code is 404" two possibilities: Either there is a value, and you can unwrap the optional to access
print("The status message is \(statusMessage)")
that value, or there isn’t a value at all.
// Prints "The status message is Not Found" NOTE
If you only need some of the tuple’s values, ignore parts of the tuple with an The concept of optionals doesn’t exist in C or Objective-C. The nearest thing in
underscore (_) when you decompose the tuple: Objective-C is the ability to return nil from a method that would otherwise return an
object, with nil meaning “the absence of a valid object.” However, this only works for
let (justTheStatusCode, _) = http404Error objects—it doesn’t work for structures, basic C types, or enumeration values. For these
print("The status code is \(justTheStatusCode)") types, Objective-C methods typically return a special value (such as NSNotFound) to
indicate the absence of a value. This approach assumes that the method’s caller knows
// Prints "The status code is 404" there is a special value to test against and remembers to check for it. Swift’s optionals
let you indicate the absence of a value for any type at all, without the need for special
Alternatively, access the individual element values in a tuple using index numbers
constants.
starting at zero:
Here’s an example of how optionals can be used to cope with the absence of a value.
print("The status code is \(http404Error.0)") Swift’s Int type has an initializer which tries to convert a String value into an Int
// Prints "The status code is 404" value. However, not every string can be converted into an integer. The string "123"
print("The status message is \(http404Error.1)") can be converted into the numeric value 123, but the string "hello, world" does not
have an obvious numeric value to convert to.
// Prints "The status message is Not Found"

You can name the individual elements in a tuple when the tuple is defined: The example below uses the initializer to try to convert a String into an Int:

let http200Status = (statusCode: 200, description: "OK") let possibleNumber = "123"

If you name the elements in a tuple, you can use the element names to access the let convertedNumber = Int(possibleNumber)
values of those elements: // convertedNumber is inferred to be of type "Int?", or "optional
Int"
print("The status code is \(http200Status.statusCode)")
Because the initializer might fail, it returns an optional Int, rather than an Int. An
// Prints "The status code is 200" optional Int is written as Int?, not Int. The question mark indicates that the value it
print("The status message is \(http200Status.description)") contains is optional, meaning that it might contain some Int value, or it might contain
// Prints "The status message is OK" no value at all. (It can’t contain anything else, such as a Bool value or a String value.
It’s either an Int, or it’s nothing at all.)
Tuples are particularly useful as the return values of functions. A function that tries to
retrieve a web page might return the (Int, String) tuple type to describe the
nil
success or failure of the page retrieval. By returning a tuple with two distinct values,
each of a different type, the function provides more useful information about its You set an optional variable to a valueless state by assigning it the special value nil:
outcome than if it could only return a single value of a single type. For more
var serverResponseCode: Int? = 404
information, see Functions with Multiple Return Values.
// serverResponseCode contains an actual Int value of 404
NOTE serverResponseCode = nil
Tuples are useful for temporary groups of related values. They are not suited to the // serverResponseCode now contains no value
creation of complex data structures. If your data structure is likely to persist beyond a NOTE

20
 nil cannot be used with nonoptional constants and variables. If a constant or variable You use optional binding to find out whether an optional contains a value, and if so, to
in your code needs to work with the absence of a value under certain conditions, make that value available as a temporary constant or variable. Optional binding can
always declare it as an optional value of the appropriate type.
be used with if and while statements to check for a value inside an optional, and to
If you define an optional variable without providing a default value, the variable is extract that value into a constant or variable, as part of a single action. if and while
automatically set to nil for you: statements are described in more detail in Control Flow.

var surveyAnswer: String? Write an optional binding for an if statement as follows:

// surveyAnswer is automatically set to nil
if let constantName = someOptional {
NOTE statements
}
Swift’s nil is not the same as nil in Objective-C. In Objective-C, nil is a pointer to a
nonexistent object. In Swift, nil is not a pointer—it is the absence of a value of a You can rewrite the possibleNumber example from the Optionals section to use
certain type. Optionals of any type can be set to nil, not just object types. optional binding rather than forced unwrapping:

If Statements and Forced Unwrapping if let actualNumber = Int(possibleNumber) {

You can use an if statement to find out whether an optional contains a value by print("\"\(possibleNumber)\" has an integer value of \
comparing the optional against nil. You perform this comparison with the “equal to” (actualNumber)")
operator (==) or the “not equal to” operator (!=). } else {
print("\"\(possibleNumber)\" could not be converted to an
If an optional has a value, it is considered to be “not equal to” nil: integer")
}
if convertedNumber != nil {
// Prints ""123" has an integer value of 123"
print("convertedNumber contains some integer value.")
This code can be read as:
}
// Prints "convertedNumber contains some integer value."
“If the optional Int returned by Int(possibleNumber) contains a value, set a new
Once you’re sure that the optional does contain a value, you can access its underlying constant called actualNumber to the value contained in the optional.”
value by adding an exclamation mark (!) to the end of the optional’s name. The
exclamation mark effectively says, “I know that this optional definitely has a value; If the conversion is successful, the actualNumber constant becomes available for use
please use it.” This is known as forced unwrapping of the optional’s value: within the first branch of the if statement. It has already been initialized with the value
contained within the optional, and so there is no need to use the ! suffix to access its
if convertedNumber != nil { value. In this example, actualNumber is simply used to print the result of the
print("convertedNumber has an integer value of \ conversion.
(convertedNumber!).")
} You can use both constants and variables with optional binding. If you wanted to
// Prints "convertedNumber has an integer value of 123." manipulate the value of actualNumber within the first branch of the if statement, you
could write if var actualNumber instead, and the value contained within the optional
For more on the if statement, see Control Flow.
would be made available as a variable rather than a constant.
NOTE
You can include as many optional bindings and Boolean conditions in a single if
Trying to use ! to access a nonexistent optional value triggers a runtime error. Always statement as you need to, separated by commas. If any of the values in the optional
make sure that an optional contains a non-nil value before using ! to force-unwrap its
bindings are nil or any Boolean condition evaluates to false, the whole if
value.
statement’s condition is considered to be false. The following if statements are
Optional Binding equivalent:

21
 if let firstNumber = Int("4"), let secondNumber = Int("42"), time it is accessed. The following example shows the difference in behavior between
firstNumber < secondNumber && secondNumber < 100 {
an optional string and an implicitly unwrapped optional string when accessing their
print("\(firstNumber) < \(secondNumber) < 100") wrapped value as an explicit String:
}
let possibleString: String? = "An optional string."
// Prints "4 < 42 < 100"
let forcedString: String = possibleString! // requires an
exclamation mark
if let firstNumber = Int("4") {
if let secondNumber = Int("42") {
let assumedString: String! = "An implicitly unwrapped optional
if firstNumber < secondNumber && secondNumber < 100 { string."
print("\(firstNumber) < \(secondNumber) < 100") let implicitString: String = assumedString // no need for an
exclamation mark
}
} You can think of an implicitly unwrapped optional as giving permission for the optional
to be unwrapped automatically whenever it is used. Rather than placing an
}
exclamation mark after the optional’s name each time you use it, you place an
// Prints "4 < 42 < 100" exclamation mark after the optional’s type when you declare it.
NOTE
NOTE
Constants and variables created with optional binding in an if statement are available
only within the body of the if statement. In contrast, the constants and variables If an implicitly unwrapped optional is nil and you try to access its wrapped value, you’ll
created with a guard statement are available in the lines of code that follow the guard trigger a runtime error. The result is exactly the same as if you place an exclamation
statement, as described in Early Exit. mark after a normal optional that does not contain a value.

Implicitly Unwrapped Optionals You can still treat an implicitly unwrapped optional like a normal optional, to check if it
As described above, optionals indicate that a constant or variable is allowed to have contains a value:
“no value”. Optionals can be checked with an if statement to see if a value exists,
if assumedString != nil {
and can be conditionally unwrapped with optional binding to access the optional’s
value if it does exist. print(assumedString)
}
Sometimes it is clear from a program’s structure that an optional will always have a // Prints "An implicitly unwrapped optional string."
value, after that value is first set. In these cases, it is useful to remove the need to
You can also use an implicitly unwrapped optional with optional binding, to check and
check and unwrap the optional’s value every time it is accessed, because it can be
unwrap its value in a single statement:
safely assumed to have a value all of the time.
if let definiteString = assumedString {
These kinds of optionals are defined as implicitly unwrapped optionals. You write an
print(definiteString)
implicitly unwrapped optional by placing an exclamation mark (String!) rather than a
question mark (String?) after the type that you want to make optional. }
// Prints "An implicitly unwrapped optional string."
Implicitly unwrapped optionals are useful when an optional’s value is confirmed to NOTE
exist immediately after the optional is first defined and can definitely be assumed to
Do not use an implicitly unwrapped optional when there is a possibility of a variable
exist at every point thereafter. The primary use of implicitly unwrapped optionals in becoming nil at a later point. Always use a normal optional type if you need to check
Swift is during class initialization, as described in Unowned References and Implicitly for a nil value during the lifetime of a variable.
Unwrapped Optional Properties.
Error Handling
An implicitly unwrapped optional is a normal optional behind the scenes, but can also
be used like a nonoptional value, without the need to unwrap the optional value each

22
You use error handling to respond to error conditions your program may encounter } catch SandwichError.missingIngredients(let ingredients) {
during execution. buyGroceries(ingredients)
}
In contrast to optionals, which can use the presence or absence of a value to
In this example, the makeASandwich() function will throw an error if no clean dishes
communicate success or failure of a function, error handling allows you to determine
are available or if any ingredients are missing. Because makeASandwich() can throw
the underlying cause of failure, and, if necessary, propagate the error to another part
an error, the function call is wrapped in a try expression. By wrapping the function call
of your program.
in a do statement, any errors that are thrown will be propagated to the provided catch
clauses.
When a function encounters an error condition, it throws an error. That function’s caller
can then catch the error and respond appropriately.
If no error is thrown, the eatASandwich() function is called. If an error is thrown and it
func canThrowAnError() throws { matches the SandwichError.outOfCleanDishes case, then the washDishes() function
will be called. If an error is thrown and it matches the
// this function may or may not throw an error
SandwichError.missingIngredients case, then the buyGroceries(_:) function is
} called with the associated [String] value captured by the catch pattern.
A function indicates that it can throw an error by including the throws keyword in its
declaration. When you call a function that can throw an error, you prepend the try Throwing, catching, and propagating errors is covered in greater detail in Error
keyword to the expression. Handling.

Swift automatically propagates errors out of their current scope until they are handled Assertions
by a catch clause. In some cases, it is simply not possible for your code to continue execution if a
particular condition is not satisfied. In these situations, you can trigger an assertion in
do {
your code to end code execution and to provide an opportunity to debug the cause of
try canThrowAnError() the absent or invalid value.
// no error was thrown
} catch { Debugging with Assertions
// an error was thrown An assertion is a runtime check that a Boolean condition definitely evaluates to true.
} Literally put, an assertion “asserts” that a condition is true. You use an assertion to
make sure that an essential condition is satisfied before executing any further code. If
A do statement creates a new containing scope, which allows errors to be propagated the condition evaluates to true, code execution continues as usual; if the condition
to one or more catch clauses. evaluates to false, code execution ends, and your app is terminated.

Here’s an example of how error handling can be used to respond to different error If your code triggers an assertion while running in a debug environment, such as when
conditions: you build and run an app in Xcode, you can see exactly where the invalid state
occurred and query the state of your app at the time that the assertion was triggered.
func makeASandwich() throws {
An assertion also lets you provide a suitable debug message as to the nature of the
// ... assert.
}
You write an assertion by calling the Swift standard library global
assert(_:_:file:line:) function. You pass this function an expression that
do {
evaluates to true or false and a message that should be displayed if the result of the
try makeASandwich()
condition is false:
eatASandwich()
} catch SandwichError.outOfCleanDishes { let age = -3
washDishes() assert(age >= 0, "A person's age cannot be less than zero")

23
 // this causes the assertion to trigger, because age is not >= 0

In this example, code execution will continue only if age >= 0 evaluates to true, that
is, if the value of age is non-negative. If the value of age is negative, as in the code
above, then age >= 0 evaluates to false, and the assertion is triggered, terminating
the application.

The assertion message can be omitted if desired, as in the following example:

assert(age >= 0)
NOTE

Assertions are disabled when your code is compiled with optimizations, such as when
building with an app target’s default Release configuration in Xcode.

When to Use Assertions

Use an assertion whenever a condition has the potential to be false, but must
definitely be true in order for your code to continue execution. Suitable scenarios for
an assertion check include:

An integer subscript index is passed to a custom subscript implementation,

but the subscript index value could be too low or too high.
A value is passed to a function, but an invalid value means that the function
cannot fulfill its task.
An optional value is currently nil, but a non-nil value is essential for
subsequent code to execute successfully.
See also Subscripts and Functions.

NOTE

Assertions cause your app to terminate and are not a substitute for designing your
code in such a way that invalid conditions are unlikely to arise. Nonetheless, in
situations where invalid conditions are possible, an assertion is an effective way to
ensure that such conditions are highlighted and noticed during development, before
your app is published.

24
 Section 2 Assignment Operator

Basic Operators
The assignment operator (a = b) initializes or updates the value of a with the value of
b:

let b = 10
var a = 5
a = b
Basic Operators
// a is now equal to 10

An operator is a special symbol or phrase that you use to check, change, or combine If the right side of the assignment is a tuple with multiple values, its elements can be
values. For example, the addition operator (+) adds two numbers, as in let i = 1 + decomposed into multiple constants or variables at once:
2, and the logical AND operator (&&) combines two Boolean values, as in if
enteredDoorCode && passedRetinaScan. let (x, y) = (1, 2)
// x is equal to 1, and y is equal to 2
Swift supports most standard C operators and improves several capabilities to
Unlike the assignment operator in C and Objective-C, the assignment operator in
eliminate common coding errors. The assignment operator (=) does not return a value,
Swift does not itself return a value. The following statement is not valid:
to prevent it from being mistakenly used when the equal to operator (==) is intended.
Arithmetic operators (+, -, *, /, % and so forth) detect and disallow value overflow, to if x = y {
avoid unexpected results when working with numbers that become larger or smaller
// This is not valid, because x = y does not return a value.
than the allowed value range of the type that stores them. You can opt in to value
overflow behavior by using Swift’s overflow operators, as described in Overflow }
Operators. This feature prevents the assignment operator (=) from being used by accident when
the equal to operator (==) is actually intended. By making if x = y invalid, Swift helps
Swift also provides two range operators (a..<b and a...b) not found in C, as a you to avoid these kinds of errors in your code.
shortcut for expressing a range of values.
Arithmetic Operators
This chapter describes the common operators in Swift. Advanced Operators covers
Swift supports the four standard arithmetic operators for all number types:
Swift’s advanced operators, and describes how to define your own custom operators
and implement the standard operators for your own custom types.
Addition (+)

Terminology Subtraction (-)

Operators are unary, binary, or ternary: Multiplication (*)
Division (/)
Unary operators operate on a single target (such as -a). Unary prefix
1 + 2 // equals 3
operators appear immediately before their target (such as !b), and unary
postfix operators appear immediately after their target (such as c!). 5 - 3 // equals 2
2 * 3 // equals 6
Binary operators operate on two targets (such as 2 + 3) and are infix
because they appear in between their two targets. 10.0 / 2.5 // equals 4.0

Ternary operators operate on three targets. Like C, Swift has only one Unlike the arithmetic operators in C and Objective-C, the Swift arithmetic operators do
ternary operator, the ternary conditional operator (a ? b : c). not allow values to overflow by default. You can opt in to value overflow behavior by
using Swift’s overflow operators (such as a &+ b). See Overflow Operators.
The values that operators affect are operands. In the expression 1 + 2, the + symbol
is a binary operator and its two operands are the values 1 and 2. The addition operator is also supported for String concatenation:

25
 "hello, " + "world" // equals "hello, world" The sign of b is ignored for negative values of b. This means that a % b and a % -b
always give the same answer.
Remainder Operator
The remainder operator (a % b) works out how many multiples of b will fit inside a and Unary Minus Operator
returns the value that is left over (known as the remainder).
The sign of a numeric value can be toggled using a prefixed -, known as the unary
minus operator:
NOTE

The remainder operator (%) is also known as a modulo operator in other languages. let three = 3
However, its behavior in Swift for negative numbers means that it is, strictly speaking, a let minusThree = -three // minusThree equals -3
remainder rather than a modulo operation.
let plusThree = -minusThree // plusThree equals 3, or "minus
Here’s how the remainder operator works. To calculate 9 % 4, you first work out how minus three"
many 4s will fit inside 9: The unary minus operator (-) is prepended directly before the value it operates on,
without any white space.

Unary Plus Operator

The unary plus operator (+) simply returns the value it operates on, without any
change:

let minusSix = -6
You can fit two 4s inside 9, and the remainder is 1 (shown in orange). let alsoMinusSix = +minusSix // alsoMinusSix equals -6

In Swift, this would be written as: Although the unary plus operator doesn’t actually do anything, you can use it to
provide symmetry in your code for positive numbers when also using the unary minus
9 % 4 // equals 1 operator for negative numbers.
To determine the answer for a % b, the % operator calculates the following equation
and returns remainder as its output: Compound Assignment Operators
Like C, Swift provides compound assignment operators that combine assignment (=)
a = (b x some multiplier) + remainder with another operation. One example is the addition assignment operator (+=):

where some multiplier is the largest number of multiples of b that will fit inside a. var a = 1
a += 2
Inserting 9 and 4 into this equation yields: // a is now equal to 3

The expression a += 2 is shorthand for a = a + 2. Effectively, the addition and the

9 = (4 x 2) + 1
assignment are combined into one operator that performs both tasks at the same
time.
The same method is applied when calculating the remainder for a negative value of a:
NOTE
-9 % 4 // equals -1
The compound assignment operators do not return a value. For example, you cannot
Inserting -9 and 4 into the equation yields: write let b = a += 2.

-9 = (4 x -2) + -1
For a complete list of the compound assignment operators provided by the Swift
standard library, see Swift Standard Library Operators Reference.
giving a remainder value of -1.
Comparison Operators

26
Swift supports all standard C comparison operators: Tuples are compared from left to right, one value at a time, until the comparison finds
two values that aren’t equal. Those two values are compared, and the result of that
Equal to (a == b) comparison determines the overall result of the tuple comparison. If all the elements
are equal, then the tuples themselves are equal. For example:
Not equal to (a != b)
Greater than (a > b) (1, "zebra") < (2, "apple") // true because 1 is less than 2;
"zebra" and "apple" are not compared
Less than (a < b)
(3, "apple") < (3, "bird") // true because 3 is equal to 3, and
Greater than or equal to (a >= b) "apple" is less than "bird"

Less than or equal to (a <= b) (4, "dog") == (4, "dog") // true because 4 is equal to 4, and
"dog" is equal to "dog"
NOTE
In the example above, you can see the left-to-right comparison behavior on the first
Swift also provides two identity operators (=== and !==), which you use to test whether line. Because 1 is less than 2, (1, "zebra") is considered less than (2, "apple"),
two object references both refer to the same object instance. For more information, see regardless of any other values in the tuples. It doesn’t matter that "zebra" isn’t less
Classes and Structures.
than "apple", because the comparison is already determined by the tuples’ first
Each of the comparison operators returns a Bool value to indicate whether or not the elements. However, when the tuples’ first elements are the same, their second
statement is true: elements are compared—this is what happens on the second and third line.

1 == 1 // true because 1 is equal to 1 NOTE

2 != 1 // true because 2 is not equal to 1 The Swift standard library includes tuple comparison operators for tuples with fewer
2 > 1 // true because 2 is greater than 1 than seven elements. To compare tuples with seven or more elements, you must
implement the comparison operators yourself.
1 < 2 // true because 1 is less than 2
1 >= 1 // true because 1 is greater than or equal to 1
Ternary Conditional Operator
2 <= 1 // false because 2 is not less than or equal to 1
The ternary conditional operator is a special operator with three parts, which takes the
Comparison operators are often used in conditional statements, such as the if form question ? answer1 : answer2. It is a shortcut for evaluating one of two
statement: expressions based on whether question is true or false. If question is true, it
evaluates answer1 and returns its value; otherwise, it evaluates answer2 and returns
let name = "world" its value.
if name == "world" {
print("hello, world") The ternary conditional operator is shorthand for the code below:
} else {
if question {
print("I'm sorry \(name), but I don't recognize you")
answer1
}
} else {
// Prints "hello, world", because name is indeed equal to "world".
answer2
For more on the if statement, see Control Flow. }

You can also compare tuples that have the same number of values, as long as each of Here’s an example, which calculates the height for a table row. The row height should
the values in the tuple can be compared. For example, both Int and String can be be 50 points taller than the content height if the row has a header, and 20 points taller
compared, which means tuples of the type (Int, String) can be compared. In if the row doesn’t have a header:
contrast, Bool can’t be compared, which means tuples that contain a Boolean value
let contentHeight = 40
can’t be compared.
let hasHeader = true
let rowHeight = contentHeight + (hasHeader ? 50 : 20)

27
 // rowHeight is equal to 90

The preceding example is shorthand for the code below: var colorNameToUse = userDefinedColorName ?? defaultColorName
// userDefinedColorName is nil, so colorNameToUse is set to the
let contentHeight = 40 default of "red"
let hasHeader = true The userDefinedColorName variable is defined as an optional String, with a default
let rowHeight: Int value of nil. Because userDefinedColorName is of an optional type, you can use the
if hasHeader { nil-coalescing operator to consider its value. In the example above, the operator is
used to determine an initial value for a String variable called colorNameToUse.
rowHeight = contentHeight + 50
Because userDefinedColorName is nil, the expression userDefinedColorName ??
} else { defaultColorName returns the value of defaultColorName, or "red".
rowHeight = contentHeight + 20
} If you assign a non-nil value to userDefinedColorName and perform the nil-
// rowHeight is equal to 90 coalescing operator check again, the value wrapped inside userDefinedColorName is
used instead of the default:
The first example’s use of the ternary conditional operator means that rowHeight can
be set to the correct value on a single line of code, which is more concise than the userDefinedColorName = "green"
code used in the second example. colorNameToUse = userDefinedColorName ?? defaultColorName
// userDefinedColorName is not nil, so colorNameToUse is set to
The ternary conditional operator provides an efficient shorthand for deciding which of "green"
two expressions to consider. Use the ternary conditional operator with care, however.
Its conciseness can lead to hard-to-read code if overused. Avoid combining multiple
instances of the ternary conditional operator into one compound statement.
Range Operators
Swift includes two range operators, which are shortcuts for expressing a range of
values.
Nil-Coalescing Operator
The nil-coalescing operator (a ?? b) unwraps an optional a if it contains a value, or Closed Range Operator
returns a default value b if a is nil. The expression a is always of an optional type.
The closed range operator (a...b) defines a range that runs from a to b, and includes
The expression b must match the type that is stored inside a.
the values a and b. The value of a must not be greater than b.
The nil-coalescing operator is shorthand for the code below:
The closed range operator is useful when iterating over a range in which you want all
a != nil ? a! : b
of the values to be used, such as with a for-in loop:

The code above uses the ternary conditional operator and forced unwrapping (a!) to for index in 1...5 {
access the value wrapped inside a when a is not nil, and to return b otherwise. The print("\(index) times 5 is \(index * 5)")
nil-coalescing operator provides a more elegant way to encapsulate this conditional
}
checking and unwrapping in a concise and readable form.
// 1 times 5 is 5
NOTE // 2 times 5 is 10
If the value of a is non-nil, the value of b is not evaluated. This is known as short- // 3 times 5 is 15
circuit evaluation. // 4 times 5 is 20
The example below uses the nil-coalescing operator to choose between a default // 5 times 5 is 25
color name and an optional user-defined color name: For more on for-in loops, see Control Flow.
let defaultColorName = "red"
Half-Open Range Operator
var userDefinedColorName: String? // defaults to nil

28
The half-open range operator (a..<b) defines a range that runs from a to b, but does // Prints "ACCESS DENIED"
not include b. It is said to be half-open because it contains its first value, but not its The phrase if !allowedEntry can be read as “if not allowed entry.” The subsequent
final value. As with the closed range operator, the value of a must not be greater than line is only executed if “not allowed entry” is true; that is, if allowedEntry is false.
b. If the value of a is equal to b, then the resulting range will be empty.
As in this example, careful choice of Boolean constant and variable names can help to
Half-open ranges are particularly useful when you work with zero-based lists such as keep code readable and concise, while avoiding double negatives or confusing logic
arrays, where it is useful to count up to (but not including) the length of the list: statements.

let names = ["Anna", "Alex", "Brian", "Jack"]

Logical AND Operator
let count = names.count
The logical AND operator (a && b) creates logical expressions where both values
for i in 0..<count { must be true for the overall expression to also be true.
print("Person \(i + 1) is called \(names[i])")
} If either value is false, the overall expression will also be false. In fact, if the first
// Person 1 is called Anna
value is false, the second value won’t even be evaluated, because it can’t possibly
make the overall expression equate to true. This is known as short-circuit evaluation.
// Person 2 is called Alex
// Person 3 is called Brian This example considers two Bool values and only allows access if both values are
// Person 4 is called Jack true:
Note that the array contains four items, but 0..<count only counts as far as 3 (the
let enteredDoorCode = true
index of the last item in the array), because it is a half-open range. For more on
arrays, see Arrays. let passedRetinaScan = false
if enteredDoorCode && passedRetinaScan {
Logical Operators print("Welcome!")
Logical operators modify or combine the Boolean logic values true and false. Swift } else {
supports the three standard logical operators found in C-based languages: print("ACCESS DENIED")
}
Logical NOT (!a)
// Prints "ACCESS DENIED"
Logical AND (a && b)
Logical OR Operator
Logical OR (a || b)
The logical OR operator (a || b) is an infix operator made from two adjacent pipe
Logical NOT Operator characters. You use it to create logical expressions in which only one of the two values
has to be true for the overall expression to be true.
The logical NOT operator (!a) inverts a Boolean value so that true becomes false,
and false becomes true.
Like the Logical AND operator above, the Logical OR operator uses short-circuit
evaluation to consider its expressions. If the left side of a Logical OR expression is
The logical NOT operator is a prefix operator, and appears immediately before the
true, the right side is not evaluated, because it cannot change the outcome of the
value it operates on, without any white space. It can be read as “not a”, as seen in the
overall expression.
following example:

let allowedEntry = false In the example below, the first Bool value (hasDoorKey) is false, but the second value
(knowsOverridePassword) is true. Because one value is true, the overall expression
if !allowedEntry {
also evaluates to true, and access is allowed:
print("ACCESS DENIED")
} let hasDoorKey = false

29
 let knowsOverridePassword = true print("ACCESS DENIED")
if hasDoorKey || knowsOverridePassword { }
print("Welcome!") // Prints "Welcome!"
} else { The parentheses make it clear that the first two values are considered as part of a
print("ACCESS DENIED") separate possible state in the overall logic. The output of the compound expression
} doesn’t change, but the overall intention is clearer to the reader. Readability is always
preferred over brevity; use parentheses where they help to make your intentions clear.
// Prints "Welcome!"

Combining Logical Operators

You can combine multiple logical operators to create longer compound expressions:

if enteredDoorCode && passedRetinaScan || hasDoorKey ||

knowsOverridePassword {
print("Welcome!")
} else {
print("ACCESS DENIED")
}
// Prints "Welcome!"

This example uses multiple && and || operators to create a longer compound
expression. However, the && and || operators still operate on only two values, so this
is actually three smaller expressions chained together. The example can be read as:

If we’ve entered the correct door code and passed the retina scan, or if we have a
valid door key, or if we know the emergency override password, then allow access.

Based on the values of enteredDoorCode, passedRetinaScan, and hasDoorKey, the

first two subexpressions are false. However, the emergency override password is
known, so the overall compound expression still evaluates to true.

NOTE

The Swift logical operators && and || are left-associative, meaning that compound
expressions with multiple logical operators evaluate the leftmost subexpression first.

Explicit Parentheses
It is sometimes useful to include parentheses when they are not strictly needed, to
make the intention of a complex expression easier to read. In the door access
example above, it is useful to add parentheses around the first part of the compound
expression to make its intent explicit:

if (enteredDoorCode && passedRetinaScan) || hasDoorKey ||

knowsOverridePassword {
print("Welcome!")
} else {

30
 Section 3 NOTE

For information about using special characters in string literals, see Special Characters

Strings and Characters in String Literals.

Initializing an Empty String

To create an empty String value as the starting point for building a longer string,
either assign an empty string literal to a variable, or initialize a new String instance
Strings and Characters with initializer syntax:

A string is a series of characters, such as "hello, world" or "albatross". Swift var emptyString = "" // empty string literal
strings are represented by the String type. The contents of a String can be var anotherEmptyString = String() // initializer syntax
accessed in various ways, including as a collection of Character values.
// these two strings are both empty, and are equivalent to each
other
Swift’s String and Character types provide a fast, Unicode-compliant way to work
with text in your code. The syntax for string creation and manipulation is lightweight Find out whether a String value is empty by checking its Boolean isEmpty property:
and readable, with a string literal syntax that is similar to C. String concatenation is as
if emptyString.isEmpty {
simple as combining two strings with the + operator, and string mutability is managed
by choosing between a constant or a variable, just like any other value in Swift. You print("Nothing to see here")
can also use strings to insert constants, variables, literals, and expressions into longer }
strings, in a process known as string interpolation. This makes it easy to create // Prints "Nothing to see here"
custom string values for display, storage, and printing.
String Mutability
Despite this simplicity of syntax, Swift’s String type is a fast, modern string
You indicate whether a particular String can be modified (or mutated) by assigning it
implementation. Every string is composed of encoding-independent Unicode
to a variable (in which case it can be modified), or to a constant (in which case it
characters, and provides support for accessing those characters in various Unicode
cannot be modified):
representations.
var variableString = "Horse"
NOTE
variableString += " and carriage"
Swift’s String type is bridged with Foundation’s NSString class. Foundation also
extends String to expose methods defined by NSString. This means, if you import // variableString is now "Horse and carriage"
Foundation, you can access those NSString methods on String without casting.
For more information about using String with Foundation and Cocoa, see Working let constantString = "Highlander"
with Cocoa Data Types in Using Swift with Cocoa and Objective-C (Swift 3.0.1). constantString += " and another Highlander"
// this reports a compile-time error - a constant string cannot be
String Literals modified
You can include predefined String values within your code as string literals. A string NOTE
literal is a fixed sequence of textual characters surrounded by a pair of double quotes This approach is different from string mutation in Objective-C and Cocoa, where you
(""). choose between two classes (NSString and NSMutableString) to indicate whether a
string can be mutated.
Use a string literal as an initial value for a constant or variable:
Strings Are Value Types
let someString = "Some string literal value"
Swift’s String type is a value type. If you create a new String value, that String
Note that Swift infers a type of String for the someString constant, because it is value is copied when it is passed to a function or method, or when it is assigned to a
initialized with a string literal value. constant or variable. In each case, a new copy of the existing String value is created,

31
and the new copy is passed or assigned, not the original version. Value types are String values can be added together (or concatenated) with the addition operator (+)
described in Structures and Enumerations Are Value Types. to create a new String value:

Swift’s copy-by-default String behavior ensures that when a function or method let string1 = "hello"
passes you a String value, it is clear that you own that exact String value, regardless let string2 = " there"
of where it came from. You can be confident that the string you are passed will not be var welcome = string1 + string2
modified unless you modify it yourself.
// welcome now equals "hello there"

Behind the scenes, Swift’s compiler optimizes string usage so that actual copying You can also append a String value to an existing String variable with the addition
takes place only when absolutely necessary. This means you always get great assignment operator (+=):
performance when working with strings as value types.
var instruction = "look over"
instruction += string2
Working with Characters
// instruction now equals "look over there"
You can access the individual Character values for a String by iterating over its
characters property with a for-in loop: You can append a Character value to a String variable with the String type’s
append() method:
for character in "Dog!🐶 ".characters {
let exclamationMark: Character = "!"
print(character)
welcome.append(exclamationMark)
}
// welcome now equals "hello there!"
// D
NOTE
// o
You can’t append a String or Character to an existing Character variable, because a
// g
Character value must contain a single character only.
// !

// 🐶 String Interpolation
The for-in loop is described in For-In Loops. String interpolation is a way to construct a new String value from a mix of constants,
variables, literals, and expressions by including their values inside a string literal. Each
Alternatively, you can create a stand-alone Character constant or variable from a item that you insert into the string literal is wrapped in a pair of parentheses, prefixed
single-character string literal by providing a Character type annotation: by a backslash:

let exclamationMark: Character = "!" let multiplier = 3

let message = "\(multiplier) times 2.5 is \(Double(multiplier) *
String values can be constructed by passing an array of Character values as an 2.5)"
argument to its initializer:
// message is "3 times 2.5 is 7.5"

let catCharacters: [Character] = ["C", "a", "t", "!", "🐱 "] In the example above, the value of multiplier is inserted into a string literal as \
(multiplier). This placeholder is replaced with the actual value of multiplier when
let catString = String(catCharacters)
the string interpolation is evaluated to create an actual string.
print(catString)

// Prints "Cat!🐱 " The value of multiplier is also part of a larger expression later in the string. This
expression calculates the value of Double(multiplier) * 2.5 and inserts the result
(7.5) into the string. In this case, the expression is written as \(Double(multiplier)
Concatenating Strings and Characters
* 2.5) when it is included inside the string literal.

32
 NOTE let sparklingHeart = "\u{1F496}" // 💖 , Unicode scalar U+1F496
The expressions you write inside parentheses within an interpolated string cannot
contain an unescaped backslash (\), a carriage return, or a line feed. However, they Extended Grapheme Clusters
can contain other string literals. Every instance of Swift’s Character type represents a single extended grapheme
cluster. An extended grapheme cluster is a sequence of one or more Unicode scalars
Unicode that (when combined) produce a single human-readable character.
Unicode is an international standard for encoding, representing, and processing text in
different writing systems. It enables you to represent almost any character from any Here’s an example. The letter é can be represented as the single Unicode scalar é
language in a standardized form, and to read and write those characters to and from (LATIN SMALL LETTER E WITH ACUTE, or U+00E9). However, the same letter can also
an external source such as a text file or web page. Swift’s String and Character be represented as a pair of scalars—a standard letter e (LATIN SMALL LETTER E, or
types are fully Unicode-compliant, as described in this section. U+0065), followed by the COMBINING ACUTE ACCENT scalar (U+0301). The COMBINING
ACUTE ACCENT scalar is graphically applied to the scalar that precedes it, turning an e
Unicode Scalars into an é when it is rendered by a Unicode-aware text-rendering system.
Behind the scenes, Swift’s native String type is built from Unicode scalar values. A
Unicode scalar is a unique 21-bit number for a character or modifier, such as U+0061 In both cases, the letter é is represented as a single Swift Character value that
for LATIN SMALL LETTER A ("a"), or U+1F425 for FRONT-FACING BABY CHICK ("🐥 "). represents an extended grapheme cluster. In the first case, the cluster contains a
single scalar; in the second case, it is a cluster of two scalars:
NOTE
let eAcute: Character = "\u{E9}" // é
A Unicode scalar is any Unicode code point in the range U+0000 to U+D7FF inclusive or let combinedEAcute: Character = "\u{65}\u{301}" // e
U+E000 to U+10FFFF inclusive. Unicode scalars do not include the Unicode surrogate followed bý
pair code points, which are the code points in the range U+D800 to U+DFFF inclusive.
// eAcute is é, combinedEAcute is é
Note that not all 21-bit Unicode scalars are assigned to a character—some scalars are
Extended grapheme clusters are a flexible way to represent many complex script
reserved for future assignment. Scalars that have been assigned to a character
characters as a single Character value. For example, Hangul syllables from the
typically also have a name, such as LATIN SMALL LETTER A and FRONT-FACING BABY
Korean alphabet can be represented as either a precomposed or decomposed
CHICK in the examples above.
sequence. Both of these representations qualify as a single Character value in Swift:
Special Characters in String Literals let precomposed: Character = "\u{D55C}" //
String literals can include the following special characters: let decomposed: Character = "\u{1112}\u{1161}\u{11AB}" // , ,

The escaped special characters \0 (null character), \\ (backslash), \t

// precomposed is , decomposed is
(horizontal tab), \n (line feed), \r (carriage return), \" (double quote) and
\' (single quote) Extended grapheme clusters enable scalars for enclosing marks (such as COMBINING
ENCLOSING CIRCLE, or U+20DD) to enclose other Unicode scalars as part of a single
An arbitrary Unicode scalar, written as \u{n}, where n is a 1–8 digit
Character value:
hexadecimal number with a value equal to a valid Unicode code point
The code below shows four examples of these special characters. The wiseWords let enclosedEAcute: Character = "\u{E9}\u{20DD}"
constant contains two escaped double quote characters. The dollarSign, // enclosedEAcute is é ⃝
blackHeart, and sparklingHeart constants demonstrate the Unicode scalar format:
Unicode scalars for regional indicator symbols can be combined in pairs to make a
let wiseWords = "\"Imagination is more important than knowledge\" - single Character value, such as this combination of REGIONAL INDICATOR SYMBOL
Einstein" LETTER U (U+1F1FA) and REGIONAL INDICATOR SYMBOL LETTER S (U+1F1F8):
// "Imagination is more important than knowledge" - Einstein
let regionalIndicatorForUS: Character = "\u{1F1FA}\u{1F1F8}"
let dollarSign = "\u{24}" // $, Unicode scalar U+0024
// regionalIndicatorForUS is &
let blackHeart = "\u{2665}" // ♥, Unicode scalar U+2665

33
Counting Characters You access and modify a string through its methods and properties, or by using
subscript syntax.
To retrieve a count of the Character values in a string, use the count property of the
string’s characters property:
String Indices
let unusualMenagerie = "Koala 🐨 , Snail 🐌 , Penguin 🐧 , Dromedary Each String value has an associated index type, String.Index, which corresponds to
the position of each Character in the string.
🐪"
print("unusualMenagerie has \(unusualMenagerie.characters.count) As mentioned above, different characters can require different amounts of memory to
characters")
store, so in order to determine which Character is at a particular position, you must
// Prints "unusualMenagerie has 40 characters" iterate over each Unicode scalar from the start or end of that String. For this reason,
Note that Swift’s use of extended grapheme clusters for Character values means that Swift strings cannot be indexed by integer values.
string concatenation and modification may not always affect a string’s character count.
Use the startIndex property to access the position of the first Character of a String.
For example, if you initialize a new string with the four-character word cafe, and then The endIndex property is the position after the last character in a String. As a result,
append a COMBINING ACUTE ACCENT (U+0301) to the end of the string, the resulting the endIndex property isn’t a valid argument to a string’s subscript. If a String is
string will still have a character count of 4, with a fourth character of é, not e: empty, startIndex and endIndex are equal.

var word = "cafe" You access the indices before and after a given index using the index(before:) and
print("the number of characters in \(word) is \ index(after:) methods of String. To access an index farther away from the given
(word.characters.count)") index, you can use the index(_:offsetBy:) method instead of calling one of these
// Prints "the number of characters in cafe is 4" methods multiple times.

You can use subscript syntax to access the Character at a particular String index.
word += "\u{301}" // COMBINING ACUTE ACCENT, U+0301

let greeting = "Guten Tag!"

print("the number of characters in \(word) is \ greeting[greeting.startIndex]
(word.characters.count)")
// G
// Prints "the number of characters in café is 4"
greeting[greeting.index(before: greeting.endIndex)]
NOTE
// !
Extended grapheme clusters can be composed of one or more Unicode scalars. This
greeting[greeting.index(after: greeting.startIndex)]
means that different characters—and different representations of the same character—
can require different amounts of memory to store. Because of this, characters in Swift // u
do not each take up the same amount of memory within a string’s representation. As a let index = greeting.index(greeting.startIndex, offsetBy: 7)
result, the number of characters in a string cannot be calculated without iterating
through the string to determine its extended grapheme cluster boundaries. If you are greeting[index]
working with particularly long string values, be aware that the characters property // a
must iterate over the Unicode scalars in the entire string in order to determine the
characters for that string. Attempting to access an index outside of a string’s range or a Character at an index
outside of a string’s range will trigger a runtime error.
The count of the characters returned by the characters property is not always the
same as the length property of an NSString that contains the same characters. The
greeting[greeting.endIndex] // Error
length of an NSString is based on the number of 16-bit code units within the string’s
UTF-16 representation and not the number of Unicode extended grapheme clusters greeting.index(after: greeting.endIndex) // Error
within the string.
Use the indices property of the characters property to access all of the indices of
individual characters in a string.
Accessing and Modifying a String

34
 for index in greeting.characters.indices { String and Character Equality
print("\(greeting[index]) ", terminator: "") String and character equality is checked with the “equal to” operator (==) and the “not
} equal to” operator (!=), as described in Comparison Operators:
// Prints "G u t e n T a g ! "
let quotation = "We're a lot alike, you and I."
NOTE
let sameQuotation = "We're a lot alike, you and I."
You can use the startIndex and endIndex properties and the index(before:),
if quotation == sameQuotation {
index(after:), and index(_:offsetBy:) methods on any type that conforms to the
Collection protocol. This includes String, as shown here, as well as collection types print("These two strings are considered equal")
such as Array, Dictionary, and Set. }

Inserting and Removing // Prints "These two strings are considered equal"

To insert a single character into a string at a specified index, use the insert(_:at:) Two String values (or two Character values) are considered equal if their extended
method, and to insert the contents of another string at a specified index, use the grapheme clusters are canonically equivalent. Extended grapheme clusters are
insert(contentsOf:at:) method. canonically equivalent if they have the same linguistic meaning and appearance, even
if they are composed from different Unicode scalars behind the scenes.
var welcome = "hello"
welcome.insert("!", at: welcome.endIndex) For example, LATIN SMALL LETTER E WITH ACUTE (U+00E9) is canonically equivalent to
LATIN SMALL LETTER E (U+0065) followed by COMBINING ACUTE ACCENT (U+0301). Both
// welcome now equals "hello!"
of these extended grapheme clusters are valid ways to represent the character é, and
so they are considered to be canonically equivalent:
welcome.insert(contentsOf:" there".characters, at:
welcome.index(before: welcome.endIndex)) // "Voulez-vous un café?" using LATIN SMALL LETTER E WITH ACUTE
// welcome now equals "hello there!" let eAcuteQuestion = "Voulez-vous un caf\u{E9}?"
To remove a single character from a string at a specified index, use the remove(at:)
method, and to remove a substring at a specified range, use the removeSubrange(_:) // "Voulez-vous un café?" using LATIN SMALL LETTER E and COMBINING
method: ACUTE ACCENT
let combinedEAcuteQuestion = "Voulez-vous un caf\u{65}\u{301}?"
welcome.remove(at: welcome.index(before: welcome.endIndex))
// welcome now equals "hello there"
if eAcuteQuestion == combinedEAcuteQuestion {
print("These two strings are considered equal")
let range = welcome.index(welcome.endIndex, offsetBy:
-6)..<welcome.endIndex }

welcome.removeSubrange(range) // Prints "These two strings are considered equal"

// welcome now equals "hello" Conversely, LATIN CAPITAL LETTER A (U+0041, or "A"), as used in English, is not
NOTE equivalent to CYRILLIC CAPITAL LETTER A (U+0410, or "А"), as used in Russian. The
characters are visually similar, but do not have the same linguistic meaning:
You can use the the insert(_:at:), insert(contentsOf:at:), remove(at:), and
removeSubrange(_:) methods on any type that conforms to the
let latinCapitalLetterA: Character = "\u{41}"
RangeReplaceableCollection protocol. This includes String, as shown here, as well
as collection types such as Array, Dictionary, and Set.
let cyrillicCapitalLetterA: Character = "\u{0410}"
Comparing Strings
Swift provides three ways to compare textual values: string and character equality, if latinCapitalLetterA != cyrillicCapitalLetterA {
prefix equality, and suffix equality. print("These two characters are not equivalent.")

35
 } var mansionCount = 0
// Prints "These two characters are not equivalent." var cellCount = 0
NOTE for scene in romeoAndJuliet {

String and character comparisons in Swift are not locale-sensitive. if scene.hasSuffix("Capulet's mansion") {
mansionCount += 1
Prefix and Suffix Equality } else if scene.hasSuffix("Friar Lawrence's cell") {
To check whether a string has a particular string prefix or suffix, call the string’s cellCount += 1
hasPrefix(_:) and hasSuffix(_:) methods, both of which take a single argument of
}
type String and return a Boolean value.
}
The examples below consider an array of strings representing the scene locations print("\(mansionCount) mansion scenes; \(cellCount) cell scenes")
from the first two acts of Shakespeare’s Romeo and Juliet: // Prints "6 mansion scenes; 2 cell scenes"
NOTE
let romeoAndJuliet = [
"Act 1 Scene 1: Verona, A public place", The hasPrefix(_:) and hasSuffix(_:) methods perform a character-by-character
canonical equivalence comparison between the extended grapheme clusters in each
"Act 1 Scene 2: Capulet's mansion", string, as described in String and Character Equality.
"Act 1 Scene 3: A room in Capulet's mansion",
"Act 1 Scene 4: A street outside Capulet's mansion", Unicode Representations of Strings
"Act 1 Scene 5: The Great Hall in Capulet's mansion", When a Unicode string is written to a text file or some other storage, the Unicode
"Act 2 Scene 1: Outside Capulet's mansion", scalars in that string are encoded in one of several Unicode-defined encoding forms.
Each form encodes the string in small chunks known as code units. These include the
"Act 2 Scene 2: Capulet's orchard",
UTF-8 encoding form (which encodes a string as 8-bit code units), the UTF-16
"Act 2 Scene 3: Outside Friar Lawrence's cell",
encoding form (which encodes a string as 16-bit code units), and the UTF-32
"Act 2 Scene 4: A street in Verona", encoding form (which encodes a string as 32-bit code units).
"Act 2 Scene 5: Capulet's mansion",
"Act 2 Scene 6: Friar Lawrence's cell" Swift provides several different ways to access Unicode representations of strings.
You can iterate over the string with a for-in statement, to access its individual
]
Character values as Unicode extended grapheme clusters. This process is described
You can use the hasPrefix(_:) method with the romeoAndJuliet array to count the in Working with Characters.
number of scenes in Act 1 of the play:
Alternatively, access a String value in one of three other Unicode-compliant
var act1SceneCount = 0
representations:
for scene in romeoAndJuliet {
if scene.hasPrefix("Act 1 ") { A collection of UTF-8 code units (accessed with the string’s utf8 property)
act1SceneCount += 1 A collection of UTF-16 code units (accessed with the string’s utf16
} property)
} A collection of 21-bit Unicode scalar values, equivalent to the string’s
print("There are \(act1SceneCount) scenes in Act 1") UTF-32 encoding form (accessed with the string’s unicodeScalars
// Prints "There are 5 scenes in Act 1" property)

Similarly, use the hasSuffix(_:) method to count the number of scenes that take Each example below shows a different representation of the following string, which is
place in or around Capulet’s mansion and Friar Lawrence’s cell: made up of the characters D, o, g, ‼ (DOUBLE EXCLAMATION MARK, or Unicode scalar
U+203C), and the 🐶 character (DOG FACE, or Unicode scalar U+1F436):

36
 let dogString = "Dog‼🐶 "

UTF-8 Representation
You can access a UTF-8 representation of a String by iterating over its utf8 property.
This property is of type String.UTF8View, which is a collection of unsigned 8-bit
(UInt8) values, one for each byte in the string’s UTF-8 representation:

for codeUnit in dogString.utf16 {

print("\(codeUnit) ", terminator: "")
}
print("")
// Prints "68 111 103 8252 55357 56374 "

Again, the first three codeUnit values (68, 111, 103) represent the characters D, o, and
g, whose UTF-16 code units have the same values as in the string’s UTF-8
for codeUnit in dogString.utf8 { representation (because these Unicode scalars represent ASCII characters).
print("\(codeUnit) ", terminator: "")
} The fourth codeUnit value (8252) is a decimal equivalent of the hexadecimal value
print("") 203C, which represents the Unicode scalar U+203C for the DOUBLE EXCLAMATION MARK
character. This character can be represented as a single code unit in UTF-16.
// 68 111 103 226 128 188 240 159 144 182

In the example above, the first three decimal codeUnit values (68, 111, 103) represent The fifth and sixth codeUnit values (55357 and 56374) are a UTF-16 surrogate pair
the characters D, o, and g, whose UTF-8 representation is the same as their ASCII representation of the DOG FACE character. These values are a high-surrogate value of
representation. The next three decimal codeUnit values (226, 128, 188) are a three- U+D83D (decimal value 55357) and a low-surrogate value of U+DC36 (decimal value
byte UTF-8 representation of the DOUBLE EXCLAMATION MARK character. The last four 56374).
codeUnit values (240, 159, 144, 182) are a four-byte UTF-8 representation of the DOG
FACE character. Unicode Scalar Representation
You can access a Unicode scalar representation of a String value by iterating over its
UTF-16 Representation unicodeScalars property. This property is of type UnicodeScalarView, which is a
You can access a UTF-16 representation of a String by iterating over its utf16 collection of values of type UnicodeScalar.
property. This property is of type String.UTF16View, which is a collection of unsigned
16-bit (UInt16) values, one for each 16-bit code unit in the string’s UTF-16 Each UnicodeScalar has a value property that returns the scalar’s 21-bit value,
representation: represented within a UInt32 value:

37
 for scalar in dogString.unicodeScalars {
print("\(scalar.value) ", terminator: "")
}
print("")
// Prints "68 111 103 8252 128054 "

The value properties for the first three UnicodeScalar values (68, 111, 103) once
again represent the characters D, o, and g.

The fourth codeUnit value (8252) is again a decimal equivalent of the hexadecimal
value 203C, which represents the Unicode scalar U+203C for the DOUBLE EXCLAMATION
MARK character.

The value property of the fifth and final UnicodeScalar, 128054, is a decimal
equivalent of the hexadecimal value 1F436, which represents the Unicode scalar
U+1F436 for the DOG FACE character.

As an alternative to querying their value properties, each UnicodeScalar value can

also be used to construct a new String value, such as with string interpolation:

for scalar in dogString.unicodeScalars {

print("\(scalar) ")
}
// D
// o
// g
// ‼

// 🐶

38
 Section 4 Arrays

Collection Types
An array stores values of the same type in an ordered list. The same value can
appear in an array multiple times at different positions.

NOTE

Swift’s Array type is bridged to Foundation’s NSArray class.

Collection Types For more information about using Array with Foundation and Cocoa, see Working with
Cocoa Data Types in Using Swift with Cocoa and Objective-C (Swift 3.0.1).
Swift provides three primary collection types, known as arrays, sets, and dictionaries,
Array Type Shorthand Syntax
for storing collections of values. Arrays are ordered collections of values. Sets are
unordered collections of unique values. Dictionaries are unordered collections of key- The type of a Swift array is written in full as Array<Element>, where Element is the
value associations. type of values the array is allowed to store. You can also write the type of an array in
shorthand form as [Element]. Although the two forms are functionally identical, the
shorthand form is preferred and is used throughout this guide when referring to the
type of an array.

Creating an Empty Array

You can create an empty array of a certain type using initializer syntax:

var someInts = [Int]()

print("someInts is of type [Int] with \(someInts.count) items.")
// Prints "someInts is of type [Int] with 0 items."
Arrays, sets, and dictionaries in Swift are always clear about the types of values and
Note that the type of the someInts variable is inferred to be [Int] from the type of the
keys that they can store. This means that you cannot insert a value of the wrong type
initializer.
into a collection by mistake. It also means you can be confident about the type of
values you will retrieve from a collection.
Alternatively, if the context already provides type information, such as a function
NOTE argument or an already typed variable or constant, you can create an empty array
with an empty array literal, which is written as [] (an empty pair of square brackets):
Swift’s array, set, and dictionary types are implemented as generic collections. For
more on generic types and collections, see Generics. someInts.append(3)
// someInts now contains 1 value of type Int
Mutability of Collections
someInts = []
If you create an array, a set, or a dictionary, and assign it to a variable, the collection
// someInts is now an empty array, but is still of type [Int]
that is created will be mutable. This means that you can change (or mutate) the
collection after it is created by adding, removing, or changing items in the collection. If Creating an Array with a Default Value
you assign an array, a set, or a dictionary to a constant, that collection is immutable,
and its size and contents cannot be changed. Swift’s Array type also provides an initializer for creating an array of a certain size
with all of its values set to the same default value. You pass this initializer a default
NOTE value of the appropriate type (called repeating): and the number of times that value is
repeated in the new array (called count):
It is good practice to create immutable collections in all cases where the collection
does not need to change. Doing so makes it easier for you to reason about your code var threeDoubles = Array(repeating: 0.0, count: 3)
and enables the Swift compiler to optimize the performance of the collections you
create. // threeDoubles is of type [Double], and equals [0.0, 0.0, 0.0]

39
Creating an Array by Adding Two Arrays Together Because all values in the array literal are of the same type, Swift can infer that
You can create a new array by adding together two existing arrays with compatible [String] is the correct type to use for the shoppingList variable.
types with the addition operator (+). The new array’s type is inferred from the type of
the two arrays you add together: Accessing and Modifying an Array
You access and modify an array through its methods and properties, or by using
var anotherThreeDoubles = Array(repeating: 2.5, count: 3) subscript syntax.
// anotherThreeDoubles is of type [Double], and equals [2.5, 2.5,
2.5] To find out the number of items in an array, check its read-only count property:

var sixDoubles = threeDoubles + anotherThreeDoubles print("The shopping list contains \(shoppingList.count) items.")

// sixDoubles is inferred as [Double], and equals [0.0, 0.0, 0.0, // Prints "The shopping list contains 2 items."
2.5, 2.5, 2.5] Use the Boolean isEmpty property as a shortcut for checking whether the count
property is equal to 0:
Creating an Array with an Array Literal
You can also initialize an array with an array literal, which is a shorthand way to write if shoppingList.isEmpty {
one or more values as an array collection. An array literal is written as a list of values, print("The shopping list is empty.")
separated by commas, surrounded by a pair of square brackets:
} else {
[value 1, value 2, value 3] print("The shopping list is not empty.")
The example below creates an array called shoppingList to store String values: }
// Prints "The shopping list is not empty."
var shoppingList: [String] = ["Eggs", "Milk"]
You can add a new item to the end of an array by calling the array’s append(_:)
// shoppingList has been initialized with two initial items
method:
The shoppingList variable is declared as “an array of string values”, written as
[String]. Because this particular array has specified a value type of String, it is shoppingList.append("Flour")
allowed to store String values only. Here, the shoppingList array is initialized with // shoppingList now contains 3 items, and someone is making
two String values ("Eggs" and "Milk"), written within an array literal. pancakes

Alternatively, append an array of one or more compatible items with the addition
NOTE
assignment operator (+=):
The shoppingList array is declared as a variable (with the var introducer) and not a
constant (with the let introducer) because more items are added to the shopping list in shoppingList += ["Baking Powder"]
the examples below.
// shoppingList now contains 4 items
In this case, the array literal contains two String values and nothing else. This shoppingList += ["Chocolate Spread", "Cheese", "Butter"]
matches the type of the shoppingList variable’s declaration (an array that can only // shoppingList now contains 7 items
contain String values), and so the assignment of the array literal is permitted as a
way to initialize shoppingList with two initial items. Retrieve a value from the array by using subscript syntax, passing the index of the
value you want to retrieve within square brackets immediately after the name of the
Thanks to Swift’s type inference, you don’t have to write the type of the array if you’re array:
initializing it with an array literal containing values of the same type. The initialization
var firstItem = shoppingList[0]
of shoppingList could have been written in a shorter form instead:
// firstItem is equal to "Eggs"
var shoppingList = ["Eggs", "Milk"] NOTE

The first item in the array has an index of 0, not 1. Arrays in Swift are always zero-
indexed.

40
You can use subscript syntax to change an existing value at a given index: // firstItem is now equal to "Six eggs"

If you want to remove the final item from an array, use the removeLast() method
shoppingList[0] = "Six eggs" rather than the remove(at:) method to avoid the need to query the array’s count
// the first item in the list is now equal to "Six eggs" rather property. Like the remove(at:) method, removeLast() returns the removed item:
than "Eggs"

You can also use subscript syntax to change a range of values at once, even if the let apples = shoppingList.removeLast()
replacement set of values has a different length than the range you are replacing. The // the last item in the array has just been removed
following example replaces "Chocolate Spread", "Cheese", and "Butter" with // shoppingList now contains 5 items, and no apples
"Bananas" and "Apples":
// the apples constant is now equal to the removed "Apples" string

shoppingList[4...6] = ["Bananas", "Apples"]

Iterating Over an Array
// shoppingList now contains 6 items
You can iterate over the entire set of values in an array with the for-in loop:
NOTE

You can’t use subscript syntax to append a new item to the end of an array. for item in shoppingList {
print(item)
To insert an item into the array at a specified index, call the array’s insert(_:at:)
}
method:
// Six eggs
shoppingList.insert("Maple Syrup", at: 0) // Milk
// shoppingList now contains 7 items // Flour
// "Maple Syrup" is now the first item in the list // Baking Powder
This call to the insert(_:at:) method inserts a new item with a value of "Maple // Bananas
Syrup" at the very beginning of the shopping list, indicated by an index of 0. If you need the integer index of each item as well as its value, use the enumerated()
method to iterate over the array instead. For each item in the array, the enumerated()
Similarly, you remove an item from the array with the remove(at:) method. This method returns a tuple composed of an integer and the item. The integers start at zero
method removes the item at the specified index and returns the removed item and count up by one for each item; if you enumerate over a whole array, these
(although you can ignore the returned value if you do not need it): integers match the items’ indices. You can decompose the tuple into temporary
constants or variables as part of the iteration:
let mapleSyrup = shoppingList.remove(at: 0)
// the item that was at index 0 has just been removed for (index, value) in shoppingList.enumerated() {
// shoppingList now contains 6 items, and no Maple Syrup print("Item \(index + 1): \(value)")
// the mapleSyrup constant is now equal to the removed "Maple }
Syrup" string
// Item 1: Six eggs
NOTE
// Item 2: Milk
If you try to access or modify a value for an index that is outside of an array’s existing // Item 3: Flour
bounds, you will trigger a runtime error. You can check that an index is valid before
using it by comparing it to the array’s count property. Except when count is 0 (meaning // Item 4: Baking Powder
the array is empty), the largest valid index in an array will always be count - 1, // Item 5: Bananas
because arrays are indexed from zero.
For more about the for-in loop, see For-In Loops.
Any gaps in an array are closed when an item is removed, and so the value at index 0
is once again equal to "Six eggs": Sets
firstItem = shoppingList[0]

41
A set stores distinct values of the same type in a collection with no defined ordering. var letters = Set<Character>()
You can use a set instead of an array when the order of items is not important, or print("letters is of type Set<Character> with \(letters.count)
when you need to ensure that an item only appears once. items.")
// Prints "letters is of type Set<Character> with 0 items."
NOTE
NOTE
Swift’s Set type is bridged to Foundation’s NSSet class. The type of the letters variable is inferred to be Set<Character>, from the type of the
For more information about using Set with Foundation and Cocoa, see Working with initializer.
Cocoa Data Types in Using Swift with Cocoa and Objective-C (Swift 3.0.1).
Alternatively, if the context already provides type information, such as a function
Hash Values for Set Types argument or an already typed variable or constant, you can create an empty set with
an empty array literal:
A type must be hashable in order to be stored in a set—that is, the type must provide
a way to compute a hash value for itself. A hash value is an Int value that is the same letters.insert("a")
for all objects that compare equally, such that if a == b, it follows that a.hashValue ==
// letters now contains 1 value of type Character
b.hashValue.
letters = []
All of Swift’s basic types (such as String, Int, Double, and Bool) are hashable by // letters is now an empty set, but is still of type Set<Character>
default, and can be used as set value types or dictionary key types. Enumeration case
values without associated values (as described in Enumerations) are also hashable by Creating a Set with an Array Literal
default. You can also initialize a set with an array literal, as a shorthand way to write one or
more values as a set collection.
NOTE

You can use your own custom types as set value types or dictionary key types by The example below creates a set called favoriteGenres to store String values:
making them conform to the Hashable protocol from Swift’s standard library. Types that
conform to the Hashable protocol must provide a gettable Int property called var favoriteGenres: Set<String> = ["Rock", "Classical", "Hip hop"]
hashValue. The value returned by a type’s hashValue property is not required to be the // favoriteGenres has been initialized with three initial items
same across different executions of the same program, or in different programs.
The favoriteGenres variable is declared as “a set of String values”, written as
Because the Hashable protocol conforms to Equatable, conforming types must also Set<String>. Because this particular set has specified a value type of String, it is
provide an implementation of the equals operator (==). The Equatable protocol
requires any conforming implementation of == to be an equivalence relation. That is, an
only allowed to store String values. Here, the favoriteGenres set is initialized with
implementation of == must satisfy the following three conditions, for all values a, b, and three String values ("Rock", "Classical", and "Hip hop"), written within an array
c: literal.
a == a (Reflexivity)
NOTE

a == b implies b == a (Symmetry) The favoriteGenres set is declared as a variable (with the var introducer) and not a
constant (with the let introducer) because items are added and removed in the
a == b && b == c implies a == c (Transitivity) examples below.

A set type cannot be inferred from an array literal alone, so the type Set must be
For more information about conforming to protocols, see Protocols.
explicitly declared. However, because of Swift’s type inference, you don’t have to write
Set Type Syntax the type of the set if you’re initializing it with an array literal containing values of the
same type. The initialization of favoriteGenres could have been written in a shorter
The type of a Swift set is written as Set<Element>, where Element is the type that the
form instead:
set is allowed to store. Unlike arrays, sets do not have an equivalent shorthand form.
var favoriteGenres: Set = ["Rock", "Classical", "Hip hop"]
Creating and Initializing an Empty Set
You can create an empty set of a certain type using initializer syntax:

42
Because all values in the array literal are of the same type, Swift can infer that // Prints "It's too funky in here."
Set<String> is the correct type to use for the favoriteGenres variable.
Iterating Over a Set
Accessing and Modifying a Set You can iterate over the values in a set with a for-in loop.
You access and modify a set through its methods and properties.
for genre in favoriteGenres {
print("\(genre)")
To find out the number of items in a set, check its read-only count property:
}
print("I have \(favoriteGenres.count) favorite music genres.") // Jazz
// Prints "I have 3 favorite music genres." // Hip hop
Use the Boolean isEmpty property as a shortcut for checking whether the count // Classical
property is equal to 0: For more about the for-in loop, see For-In Loops.
if favoriteGenres.isEmpty {
Swift’s Set type does not have a defined ordering. To iterate over the values of a set in
print("As far as music goes, I'm not picky.")
a specific order, use the sorted() method, which returns the set’s elements as an
} else { array sorted using the < operator.
print("I have particular music preferences.")
for genre in favoriteGenres.sorted() {
}
print("\(genre)")
// Prints "I have particular music preferences."
}
You can add a new item into a set by calling the set’s insert(_:) method:
// Classical
favoriteGenres.insert("Jazz") // Hip hop
// favoriteGenres now contains 4 items // Jazz

You can remove an item from a set by calling the set’s remove(_:) method, which
removes the item if it’s a member of the set, and returns the removed value, or returns Performing Set Operations
nil if the set did not contain it. Alternatively, all items in a set can be removed with its You can efficiently perform fundamental set operations, such as combining two sets
removeAll() method. together, determining which values two sets have in common, or determining whether
two sets contain all, some, or none of the same values.
if let removedGenre = favoriteGenres.remove("Rock") {
print("\(removedGenre)? I'm over it.") Fundamental Set Operations
} else { The illustration below depicts two sets—a and b—with the results of various set
print("I never much cared for that.") operations represented by the shaded regions.
}
// Prints "Rock? I'm over it."

To check whether a set contains a particular item, use the contains(_:) method.

if favoriteGenres.contains("Funk") {
print("I get up on the good foot.")
} else {
print("It's too funky in here.")
}

43
 in b are also contained by a. Set b and set c are disjoint with one another, because
they share no elements in common.

Use the intersection(_:) method to create a new set with only the values
common to both sets. Use the “is equal” operator (==) to determine whether two sets contain all of
the same values.
Use the symmetricDifference(_:) method to create a new set with values
in either set, but not both. Use the isSubset(of:) method to determine whether all of the values of a
set are contained in the specified set.
Use the union(_:) method to create a new set with all of the values in both
sets. Use the isSuperset(of:) method to determine whether a set contains all
of the values in a specified set.
Use the subtracting(_:) method to create a new set with values not in
the specified set. Use the isStrictSubset(of:) or isStrictSuperset(of:) methods to
determine whether a set is a subset or superset, but not equal to, a
let oddDigits: Set = [1, 3, 5, 7, 9]
specified set.
let evenDigits: Set = [0, 2, 4, 6, 8]
Use the isDisjoint(with:) method to determine whether two sets have
let singleDigitPrimeNumbers: Set = [2, 3, 5, 7]
any values in common.
let houseAnimals: Set = ["🐶 ", "🐱 "]
oddDigits.union(evenDigits).sorted()
// [0, 1, 2, 3, 4, 5, 6, 7, 8, 9] let farmAnimals: Set = ["🐮 ", "🐔 ", "🐑 ", "🐶 ", "🐱 "]

oddDigits.intersection(evenDigits).sorted() let cityAnimals: Set = ["🐦 ", "🐭 "]

// []
oddDigits.subtracting(singleDigitPrimeNumbers).sorted() houseAnimals.isSubset(of: farmAnimals)
// [1, 9] // true
oddDigits.symmetricDifference(singleDigitPrimeNumbers).sorted() farmAnimals.isSuperset(of: houseAnimals)
// [1, 2, 9] // true

Set Membership and Equality farmAnimals.isDisjoint(with: cityAnimals)

// true
The illustration below depicts three sets—a, b and c—with overlapping regions
representing elements shared among sets. Set a is a superset of set b, because a
contains all elements in b. Conversely, set b is a subset of set a, because all elements Dictionaries

44
A dictionary stores associations between keys of the same type and values of the Creating a Dictionary with a Dictionary Literal
same type in a collection with no defined ordering. Each value is associated with a You can also initialize a dictionary with a dictionary literal, which has a similar syntax
unique key, which acts as an identifier for that value within the dictionary. Unlike items to the array literal seen earlier. A dictionary literal is a shorthand way to write one or
in an array, items in a dictionary do not have a specified order. You use a dictionary more key-value pairs as a Dictionary collection.
when you need to look up values based on their identifier, in much the same way that
a real-world dictionary is used to look up the definition for a particular word. A key-value pair is a combination of a key and a value. In a dictionary literal, the key
and value in each key-value pair are separated by a colon. The key-value pairs are
NOTE
written as a list, separated by commas, surrounded by a pair of square brackets:
Swift’s Dictionary type is bridged to Foundation’s NSDictionary class.
[key 1: value 1, key 2: value 2, key 3: value 3]
For more information about using Dictionary with Foundation and Cocoa, see
Working with Cocoa Data Types in Using Swift with Cocoa and Objective-C (Swift The example below creates a dictionary to store the names of international airports. In
3.0.1). this dictionary, the keys are three-letter International Air Transport Association codes,
and the values are airport names:
Dictionary Type Shorthand Syntax
The type of a Swift dictionary is written in full as Dictionary<Key, Value>, where Key var airports: [String: String] = ["YYZ": "Toronto Pearson", "DUB":
"Dublin"]
is the type of value that can be used as a dictionary key, and Value is the type of value
that the dictionary stores for those keys. The airports dictionary is declared as having a type of [String: String], which
means “a Dictionary whose keys are of type String, and whose values are also of
NOTE type String”.
A dictionary Key type must conform to the Hashable protocol, like a set’s value type.
NOTE
You can also write the type of a dictionary in shorthand form as [Key: Value]. The airports dictionary is declared as a variable (with the var introducer), and not a
Although the two forms are functionally identical, the shorthand form is preferred and constant (with the let introducer), because more airports are added to the dictionary in
is used throughout this guide when referring to the type of a dictionary. the examples below.

The airports dictionary is initialized with a dictionary literal containing two key-value
Creating an Empty Dictionary
pairs. The first pair has a key of "YYZ" and a value of "Toronto Pearson". The second
As with arrays, you can create an empty Dictionary of a certain type by using pair has a key of "DUB" and a value of "Dublin".
initializer syntax:
This dictionary literal contains two String: String pairs. This key-value type matches
var namesOfIntegers = [Int: String]()
the type of the airports variable declaration (a dictionary with only String keys, and
// namesOfIntegers is an empty [Int: String] dictionary only String values), and so the assignment of the dictionary literal is permitted as a
This example creates an empty dictionary of type [Int: String] to store human- way to initialize the airports dictionary with two initial items.
readable names of integer values. Its keys are of type Int, and its values are of type
String. As with arrays, you don’t have to write the type of the dictionary if you’re initializing it
with a dictionary literal whose keys and values have consistent types. The initialization
If the context already provides type information, you can create an empty dictionary of airports could have been written in a shorter form instead:
with an empty dictionary literal, which is written as [:] (a colon inside a pair of square
brackets): var airports = ["YYZ": "Toronto Pearson", "DUB": "Dublin"]

Because all keys in the literal are of the same type as each other, and likewise all
namesOfIntegers[16] = "sixteen" values are of the same type as each other, Swift can infer that [String: String] is
// namesOfIntegers now contains 1 key-value pair the correct type to use for the airports dictionary.
namesOfIntegers = [:]
// namesOfIntegers is once again an empty dictionary of type [Int: Accessing and Modifying a Dictionary
String]

45
You access and modify a dictionary through its methods and properties, or by using }
subscript syntax. // Prints "The old value for DUB was Dublin."

You can also use subscript syntax to retrieve a value from the dictionary for a
As with an array, you find out the number of items in a Dictionary by checking its particular key. Because it is possible to request a key for which no value exists, a
read-only count property: dictionary’s subscript returns an optional value of the dictionary’s value type. If the
dictionary contains a value for the requested key, the subscript returns an optional
print("The airports dictionary contains \(airports.count) items.")
value containing the existing value for that key. Otherwise, the subscript returns nil:
// Prints "The airports dictionary contains 2 items."

Use the Boolean isEmpty property as a shortcut for checking whether the count if let airportName = airports["DUB"] {
property is equal to 0: print("The name of the airport is \(airportName).")
} else {
if airports.isEmpty {
print("That airport is not in the airports dictionary.")
print("The airports dictionary is empty.")
}
} else {
// Prints "The name of the airport is Dublin Airport."
print("The airports dictionary is not empty.")
You can use subscript syntax to remove a key-value pair from a dictionary by
}
assigning a value of nil for that key:
// Prints "The airports dictionary is not empty."

You can add a new item to a dictionary with subscript syntax. Use a new key of the airports["APL"] = "Apple International"
appropriate type as the subscript index, and assign a new value of the appropriate // "Apple International" is not the real airport for APL, so delete
type: it
airports["APL"] = nil
airports["LHR"] = "London" // APL has now been removed from the dictionary
// the airports dictionary now contains 3 items
Alternatively, remove a key-value pair from a dictionary with the
You can also use subscript syntax to change the value associated with a particular removeValue(forKey:) method. This method removes the key-value pair if it exists
key: and returns the removed value, or returns nil if no value existed:

airports["LHR"] = "London Heathrow" if let removedValue = airports.removeValue(forKey: "DUB") {

// the value for "LHR" has been changed to "London Heathrow" print("The removed airport's name is \(removedValue).")
As an alternative to subscripting, use a dictionary’s updateValue(_:forKey:) method } else {
to set or update the value for a particular key. Like the subscript examples above, the print("The airports dictionary does not contain a value for
updateValue(_:forKey:) method sets a value for a key if none exists, or updates the DUB.")
value if that key already exists. Unlike a subscript, however, the }
updateValue(_:forKey:) method returns the old value after performing an update. // Prints "The removed airport's name is Dublin Airport."
This enables you to check whether or not an update took place.
Iterating Over a Dictionary
The updateValue(_:forKey:) method returns an optional value of the dictionary’s You can iterate over the key-value pairs in a dictionary with a for-in loop. Each item
value type. For a dictionary that stores String values, for example, the method returns in the dictionary is returned as a (key, value) tuple, and you can decompose the
a value of type String?, or “optional String”. This optional value contains the old tuple’s members into temporary constants or variables as part of the iteration:
value for that key if one existed before the update, or nil if no value existed:
for (airportCode, airportName) in airports {
if let oldValue = airports.updateValue("Dublin Airport", forKey:
"DUB") { print("\(airportCode): \(airportName)")

print("The old value for DUB was \(oldValue).") }

46
 // YYZ: Toronto Pearson
// LHR: London Heathrow

For more about the for-in loop, see For-In Loops.

You can also retrieve an iterable collection of a dictionary’s keys or values by

accessing its keys and values properties:

for airportCode in airports.keys {

print("Airport code: \(airportCode)")
}
// Airport code: YYZ
// Airport code: LHR

for airportName in airports.values {

print("Airport name: \(airportName)")
}
// Airport name: Toronto Pearson
// Airport name: London Heathrow

If you need to use a dictionary’s keys or values with an API that takes an Array
instance, initialize a new array with the keys or values property:

let airportCodes = [String](airports.keys)

// airportCodes is ["YYZ", "LHR"]

let airportNames = [String](airports.values)

// airportNames is ["Toronto Pearson", "London Heathrow"]

Swift’s Dictionary type does not have a defined ordering. To iterate over the keys or
values of a dictionary in a specific order, use the sorted() method on its keys or
values property.

47
 Section 5 times table for the current value of index. After the statement is executed, the value of
index is updated to contain the second value in the range (2), and the

Control Flow print(_:separator:terminator:) function is called again. This process continues

until the end of the range is reached.

In the example above, index is a constant whose value is automatically set at the
start of each iteration of the loop. As such, index does not have to be declared before
Control Flow it is used. It is implicitly declared simply by its inclusion in the loop declaration, without
the need for a let declaration keyword.
Swift provides a variety of control flow statements. These include while loops to
perform a task multiple times; if, guard, and switch statements to execute different If you don’t need each value from a sequence, you can ignore the values by using an
branches of code based on certain conditions; and statements such as break and underscore in place of a variable name.
continue to transfer the flow of execution to another point in your code.
let base = 3

Swift also provides a for-in loop that makes it easy to iterate over arrays, let power = 10
dictionaries, ranges, strings, and other sequences. var answer = 1
for _ in 1...power {
Swift’s switch statement is also considerably more powerful than its counterpart in answer *= base
many C-like languages. Because the cases of a switch statement do not fall through
}
to the next case in Swift, it avoids common C errors caused by missing break
statements. Cases can match many different patterns, including interval matches, print("\(base) to the power of \(power) is \(answer)")
tuples, and casts to a specific type. Matched values in a switch case can be bound to // Prints "3 to the power of 10 is 59049"
temporary constants or variables for use within the case’s body, and complex
The example above calculates the value of one number to the power of another (in
matching conditions can be expressed with a where clause for each case.
this case, 3 to the power of 10). It multiplies a starting value of 1 (that is, 3 to the
power of 0) by 3, ten times, using a closed range that starts with 1 and ends with 10.
For-In Loops For this calculation, the individual counter values each time through the loop are
You use the for-in loop to iterate over a sequence, such as ranges of numbers, items unnecessary—the code simply executes the loop the correct number of times. The
in an array, or characters in a string. underscore character (_) used in place of a loop variable causes the individual values
to be ignored and does not provide access to the current value during each iteration
This example prints the first few entries in the five-times table: of the loop.

for index in 1...5 { Use a for-in loop with an array to iterate over its items.
print("\(index) times 5 is \(index * 5)")
let names = ["Anna", "Alex", "Brian", "Jack"]
}
for name in names {
// 1 times 5 is 5
print("Hello, \(name)!")
// 2 times 5 is 10
}
// 3 times 5 is 15
// Hello, Anna!
// 4 times 5 is 20
// Hello, Alex!
// 5 times 5 is 25
// Hello, Brian!
The sequence being iterated over is a range of numbers from 1 to 5, inclusive, as
// Hello, Jack!
indicated by the use of the closed range operator (...). The value of index is set to
the first number in the range (1), and the statements inside the loop are executed. In You can also iterate over a dictionary to access its key-value pairs. Each item in the
this case, the loop contains only one statement, which prints an entry from the five- dictionary is returned as a (key, value) tuple when the dictionary is iterated, and you

48
can decompose the (key, value) tuple’s members as explicitly named constants for
use within the body of the for-in loop. Here, the dictionary’s keys are decomposed
into a constant called animalName, and the dictionary’s values are decomposed into a
constant called legCount.

let numberOfLegs = ["spider": 8, "ant": 6, "cat": 4]

for (animalName, legCount) in numberOfLegs {
print("\(animalName)s have \(legCount) legs")
}
// ants have 6 legs
// spiders have 8 legs The rules of the game are as follows:
// cats have 4 legs
The board has 25 squares, and the aim is to land on or beyond square 25.
Items in a Dictionary may not necessarily be iterated in the same order in which they
were inserted. The contents of a Dictionary are inherently unordered, and iterating Each turn, you roll a six-sided dice and move by that number of squares,
over them does not guarantee the order in which they will be retrieved. For more on following the horizontal path indicated by the dotted arrow above.
arrays and dictionaries, see Collection Types. If your turn ends at the bottom of a ladder, you move up that ladder.
If your turn ends at the head of a snake, you move down that snake.
While Loops
The game board is represented by an array of Int values. Its size is based on a
A while loop performs a set of statements until a condition becomes false. These
constant called finalSquare, which is used to initialize the array and also to check for
kinds of loops are best used when the number of iterations is not known before the
a win condition later in the example. The board is initialized with 26 zero Int values,
first iteration begins. Swift provides two kinds of while loops:
not 25 (one each at indexes 0 through 25).

while evaluates its condition at the start of each pass through the loop. let finalSquare = 25
repeat-while evaluates its condition at the end of each pass through the var board = [Int](repeating: 0, count: finalSquare + 1)
loop.
Some squares are then set to have more specific values for the snakes and ladders.
While Squares with a ladder base have a positive number to move you up the board,
whereas squares with a snake head have a negative number to move you back down
A while loop starts by evaluating a single condition. If the condition is true, a set of the board.
statements is repeated until the condition becomes false.
board[03] = +08; board[06] = +11; board[09] = +09; board[10] = +02
Here’s the general form of a while loop: board[14] = -10; board[19] = -11; board[22] = -02; board[24] = -08

while condition { Square 3 contains the bottom of a ladder that moves you up to square 11. To
statements represent this, board[03] is equal to +08, which is equivalent to an integer value of 8
}
(the difference between 3 and 11). The unary plus operator (+i) balances with the
This example plays a simple game of Snakes and Ladders (also known as Chutes unary minus operator (-i), and numbers lower than 10 are padded with zeros so that
and Ladders): all board definitions align. (Neither stylistic tweak is strictly necessary, but they lead to
neater code.)

The player’s starting square is “square zero”, which is just off the bottom-left corner of
the board. The first dice roll always moves the player onto the board.

var square = 0

49
 var diceRoll = 0 The other variation of the while loop, known as the repeat-while loop, performs a
while square < finalSquare { single pass through the loop block first, before considering the loop’s condition. It then
// roll the dice continues to repeat the loop until the condition is false.
diceRoll += 1
NOTE
if diceRoll == 7 { diceRoll = 1 }
The repeat-while loop in Swift is analogous to a do-while loop in other languages.
// move by the rolled amount
square += diceRoll Here’s the general form of a repeat-while loop:
if square < board.count {
repeat {
// if we're still on the board, move up or down for a snake statements
or a ladder } while condition
square += board[square] Here’s the Snakes and Ladders example again, written as a repeat-while loop rather
} than a while loop. The values of finalSquare, board, square, and diceRoll are
initialized in exactly the same way as with a while loop.
}
print("Game over!") let finalSquare = 25
The example above uses a very simple approach to dice rolling. Instead of generating var board = [Int](repeating: 0, count: finalSquare + 1)
a random number, it starts with a diceRoll value of 0. Each time through the while board[03] = +08; board[06] = +11; board[09] = +09; board[10] = +02
loop, diceRoll is incremented by one and is then checked to see whether it has
board[14] = -10; board[19] = -11; board[22] = -02; board[24] = -08
become too large. Whenever this return value equals 7, the dice roll has become too
large and is reset to a value of 1. The result is a sequence of diceRoll values that is var square = 0
always 1, 2, 3, 4, 5, 6, 1, 2 and so on. var diceRoll = 0

In this version of the game, the first action in the loop is to check for a ladder or a
After rolling the dice, the player moves forward by diceRoll squares. It’s possible that snake. No ladder on the board takes the player straight to square 25, and so it isn’t
the dice roll may have moved the player beyond square 25, in which case the game is possible to win the game by moving up a ladder. Therefore, it’s safe to check for a
over. To cope with this scenario, the code checks that square is less than the board snake or a ladder as the first action in the loop.
array’s count property before adding the value stored in board[square] onto the
current square value to move the player up or down any ladders or snakes. At the start of the game, the player is on “square zero”. board[0] always equals 0 and
has no effect.
NOTE

Had this check not been performed, board[square] might try to access a value repeat {
outside the bounds of the board array, which would trigger an error. If square were // move up or down for a snake or ladder
equal to 26, the code would try to check the value of board[26], which is larger than
the size of the array. square += board[square]
// roll the dice
The current while loop execution then ends, and the loop’s condition is checked to
diceRoll += 1
see if the loop should be executed again. If the player has moved on or beyond
square number 25, the loop’s condition evaluates to false and the game ends. if diceRoll == 7 { diceRoll = 1 }
// move by the rolled amount
A while loop is appropriate in this case, because the length of the game is not clear at square += diceRoll
the start of the while loop. Instead, the loop is executed until a particular condition is } while square < finalSquare
satisfied.
print("Game over!")

Repeat-While After the code checks for snakes and ladders, the dice is rolled and the player is
moved forward by diceRoll squares. The current loop execution then ends.

50
The loop’s condition (while square < finalSquare) is the same as before, but this print("It's not that cold. Wear a t-shirt.")
time it’s not evaluated until the end of the first run through the loop. The structure of }
the repeat-while loop is better suited to this game than the while loop in the previous // Prints "It's not that cold. Wear a t-shirt."
example. In the repeat-while loop above, square += board[square] is always
executed immediately after the loop’s while condition confirms that square is still on One of these two branches is always executed. Because the temperature has
the board. This behavior removes the need for the array bounds check seen in the increased to 40 degrees Fahrenheit, it is no longer cold enough to advise wearing a
earlier version of the game. scarf and so the else branch is triggered instead.

You can chain multiple if statements together to consider additional clauses.

Conditional Statements
It is often useful to execute different pieces of code based on certain conditions. You temperatureInFahrenheit = 90
might want to run an extra piece of code when an error occurs, or to display a if temperatureInFahrenheit <= 32 {
message when a value becomes too high or too low. To do this, you make parts of
print("It's very cold. Consider wearing a scarf.")
your code conditional.
} else if temperatureInFahrenheit >= 86 {
Swift provides two ways to add conditional branches to your code: the if statement print("It's really warm. Don't forget to wear sunscreen.")
and the switch statement. Typically, you use the if statement to evaluate simple } else {
conditions with only a few possible outcomes. The switch statement is better suited to print("It's not that cold. Wear a t-shirt.")
more complex conditions with multiple possible permutations and is useful in
}
situations where pattern matching can help select an appropriate code branch to
execute. // Prints "It's really warm. Don't forget to wear sunscreen."

Here, an additional if statement was added to respond to particularly warm

If temperatures. The final else clause remains, and it prints a response for any
In its simplest form, the if statement has a single if condition. It executes a set of temperatures that are neither too warm nor too cold.
statements only if that condition is true.
The final else clause is optional, however, and can be excluded if the set of conditions
var temperatureInFahrenheit = 30 does not need to be complete.
if temperatureInFahrenheit <= 32 {
temperatureInFahrenheit = 72
print("It's very cold. Consider wearing a scarf.")
if temperatureInFahrenheit <= 32 {
}
print("It's very cold. Consider wearing a scarf.")
// Prints "It's very cold. Consider wearing a scarf."
} else if temperatureInFahrenheit >= 86 {
The example above checks whether the temperature is less than or equal to 32
print("It's really warm. Don't forget to wear sunscreen.")
degrees Fahrenheit (the freezing point of water). If it is, a message is printed.
Otherwise, no message is printed, and code execution continues after the if }
statement’s closing brace. Because the temperature is neither too cold nor too warm to trigger the if or else if
conditions, no message is printed.
The if statement can provide an alternative set of statements, known as an else
clause, for situations when the if condition is false. These statements are indicated Switch
by the else keyword. A switch statement considers a value and compares it against several possible
matching patterns. It then executes an appropriate block of code, based on the first
temperatureInFahrenheit = 40
pattern that matches successfully. A switch statement provides an alternative to the
if temperatureInFahrenheit <= 32 { if statement for responding to multiple potential states.
print("It's very cold. Consider wearing a scarf.")
} else {

51
In its simplest form, a switch statement compares a value against one or more values No Implicit Fallthrough
of the same type. In contrast with switch statements in C and Objective-C, switch statements in Swift
do not fall through the bottom of each case and into the next one by default. Instead,
switch some value to consider { the entire switch statement finishes its execution as soon as the first matching switch
case value 1:
respond to value 1 case is completed, without requiring an explicit break statement. This makes the
case value 2, switch statement safer and easier to use than the one in C and avoids executing
value 3: more than one switch case by mistake.
respond to value 2 or 3
default:
otherwise, do something else NOTE
}
Although break is not required in Swift, you can use a break statement to match and
Every switch statement consists of multiple possible cases, each of which begins with ignore a particular case or to break out of a matched case before that case has
the case keyword. In addition to comparing against specific values, Swift provides completed its execution. For details, see Break in a Switch Statement.
several ways for each case to specify more complex matching patterns. These options
The body of each case must contain at least one executable statement. It is not valid
are described later in this chapter.
to write the following code, because the first case is empty:

Like the body of an if statement, each case is a separate branch of code execution. let anotherCharacter: Character = "a"
The switch statement determines which branch should be selected. This procedure is
switch anotherCharacter {
known as switching on the value that is being considered.
case "a": // Invalid, the case has an empty body

Every switch statement must be exhaustive. That is, every possible value of the type case "A":
being considered must be matched by one of the switch cases. If it’s not appropriate print("The letter A")
to provide a case for every possible value, you can define a default case to cover any default:
values that are not addressed explicitly. This default case is indicated by the default
print("Not the letter A")
keyword, and must always appear last.
}

This example uses a switch statement to consider a single lowercase character called // This will report a compile-time error.
someCharacter: Unlike a switch statement in C, this switch statement does not match both "a" and
"A". Rather, it reports a compile-time error that case "a": does not contain any
let someCharacter: Character = "z" executable statements. This approach avoids accidental fallthrough from one case to
switch someCharacter { another and makes for safer code that is clearer in its intent.
case "a":
print("The first letter of the alphabet") To make a switch with a single case that matches both "a" and "A", combine the two
values into a compound case, separating the values with commas.
case "z":
print("The last letter of the alphabet") let anotherCharacter: Character = "a"
default: switch anotherCharacter {
print("Some other character") case "a", "A":
} print("The letter A")
// Prints "The last letter of the alphabet" default:
The switch statement’s first case matches the first letter of the English alphabet, a, print("Not the letter A")
and its second case matches the last letter, z. Because the switch must have a case }
for every possible character, not just every alphabetic character, this switch statement
// Prints "The letter A"
uses a default case to match all characters other than a and z. This provision
ensures that the switch statement is exhaustive.

52
For readability, a compound case can also be written over multiple lines. For more The example below takes an (x, y) point, expressed as a simple tuple of type (Int,
information about compound cases, see Compound Cases. Int), and categorizes it on the graph that follows the example.

NOTE let somePoint = (1, 1)

To explicitly fall through at the end of a particular switch case, use the fallthrough switch somePoint {
keyword, as described in Fallthrough. case (0, 0):
print("(0, 0) is at the origin")
Interval Matching
case (_, 0):
Values in switch cases can be checked for their inclusion in an interval. This example
uses number intervals to provide a natural-language count for numbers of any size: print("(\(somePoint.0), 0) is on the x-axis")
case (0, _):
let approximateCount = 62 print("(0, \(somePoint.1)) is on the y-axis")
let countedThings = "moons orbiting Saturn" case (-2...2, -2...2):
var naturalCount: String print("(\(somePoint.0), \(somePoint.1)) is inside the box")
switch approximateCount { default:
case 0: print("(\(somePoint.0), \(somePoint.1)) is outside of the box")
naturalCount = "no" }
case 1..<5: // Prints "(1, 1) is inside the box"
naturalCount = "a few"
case 5..<12:
naturalCount = "several"
case 12..<100:
naturalCount = "dozens of"
case 100..<1000:
naturalCount = "hundreds of"
default:
naturalCount = "many"
}
print("There are \(naturalCount) \(countedThings).")
// Prints "There are dozens of moons orbiting Saturn."

In the above example, approximateCount is evaluated in a switch statement. Each

case compares that value to a number or interval. Because the value of
approximateCount falls between 12 and 100, naturalCount is assigned the value
"dozens of", and execution is transferred out of the switch statement.

Tuples
You can use tuples to test multiple values in the same switch statement. Each
element of the tuple can be tested against a different value or interval of values. The switch statement determines whether the point is at the origin (0, 0), on the red x-
Alternatively, use the underscore character (_), also known as the wildcard pattern, to axis, on the orange y-axis, inside the blue 4-by-4 box centered on the origin, or
match any possible value. outside of the box.

53
Unlike C, Swift allows multiple switch cases to consider the same value or values. In
fact, the point (0, 0) could match all four of the cases in this example. However, if
multiple matches are possible, the first matching case is always used. The point (0, 0)
would match case (0, 0) first, and so all other matching cases would be ignored.

Value Bindings
A switch case can bind the value or values it matches to temporary constants or
variables, for use in the body of the case. This behavior is known as value binding,
because the values are bound to temporary constants or variables within the case’s
body.

The example below takes an (x, y) point, expressed as a tuple of type (Int, Int),
and categorizes it on the graph that follows:

let anotherPoint = (2, 0)

switch anotherPoint {
case (let x, 0):
print("on the x-axis with an x value of \(x)")
case (0, let y):
print("on the y-axis with a y value of \(y)") The switch statement determines whether the point is on the red x-axis, on the
case let (x, y): orange y-axis, or elsewhere (on neither axis).
print("somewhere else at (\(x), \(y))")
The three switch cases declare placeholder constants x and y, which temporarily take
}
on one or both tuple values from anotherPoint. The first case, case (let x, 0),
// Prints "on the x-axis with an x value of 2" matches any point with a y value of 0 and assigns the point’s x value to the temporary
constant x. Similarly, the second case, case (0, let y), matches any point with an x
value of 0 and assigns the point’s y value to the temporary constant y.

After the temporary constants are declared, they can be used within the case’s code
block. Here, they are used to print the categorization of the point.

This switch statement does not have a default case. The final case, case let (x,
y), declares a tuple of two placeholder constants that can match any value. Because
anotherPoint is always a tuple of two values, this case matches all possible
remaining values, and a default case is not needed to make the switch statement
exhaustive.

Where
A switch case can use a where clause to check for additional conditions.

The example below categorizes an (x, y) point on the following graph:

let yetAnotherPoint = (1, -1)

switch yetAnotherPoint {

54
 case let (x, y) where x == y: match, then the case is considered to match. The patterns can be written over multiple
print("(\(x), \(y)) is on the line x == y") lines if the list is long. For example:
case let (x, y) where x == -y:
let someCharacter: Character = "e"
print("(\(x), \(y)) is on the line x == -y")
switch someCharacter {
case let (x, y):
case "a", "e", "i", "o", "u":
print("(\(x), \(y)) is just some arbitrary point")
print("\(someCharacter) is a vowel")
}
case "b", "c", "d", "f", "g", "h", "j", "k", "l", "m",
// Prints "(1, -1) is on the line x == -y"
"n", "p", "q", "r", "s", "t", "v", "w", "x", "y", "z":
print("\(someCharacter) is a consonant")
default:
print("\(someCharacter) is not a vowel or a consonant")
}
// Prints "e is a vowel"

The switch statement’s first case matches all five lowercase vowels in the English
language. Similarly, its second case matches all lowercase English consonants.
Finally, the default case matches any other character.

Compound cases can also include value bindings. All of the patterns of a compound
case have to include the same set of value bindings, and each binding has to get a
value of the same type from all of the patterns in the compound case. This ensures
that, no matter which part of the compound case matched, the code in the body of the
case can always access a value for the bindings and that the value always has the
same type.

let stillAnotherPoint = (9, 0)

switch stillAnotherPoint {
case (let distance, 0), (0, let distance):
print("On an axis, \(distance) from the origin")
The switch statement determines whether the point is on the green diagonal line
where x == y, on the purple diagonal line where x == -y, or neither. default:
print("Not on an axis")
The three switch cases declare placeholder constants x and y, which temporarily take }
on the two tuple values from yetAnotherPoint. These constants are used as part of a // Prints "On an axis, 9 from the origin"
where clause, to create a dynamic filter. The switch case matches the current value of
point only if the where clause’s condition evaluates to true for that value. The case above has two patterns: (let distance, 0) matches points on the x-axis
and (0, let distance) matches points on the y-axis. Both patterns include a binding
As in the previous example, the final case matches all possible remaining values, and for distance and distance is an integer in both patterns—which means that the code
so a default case is not needed to make the switch statement exhaustive. in the body of the case can always access a value for distance.

Compound Cases Control Transfer Statements

Multiple switch cases that share the same body can be combined by writing several
patterns after case, with a comma between each of the patterns. If any of the patterns

55
Control transfer statements change the order in which your code is executed, by you want to terminate the execution of the switch or loop statement earlier than would
transferring control from one piece of code to another. Swift has five control transfer otherwise be the case.
statements:
Break in a Loop Statement
continue
When used inside a loop statement, break ends the loop’s execution immediately and
break transfers control to the code after the loop’s closing brace (}). No further code from
fallthrough the current iteration of the loop is executed, and no further iterations of the loop are
return started.
throw
Break in a Switch Statement
The continue, break, and fallthrough statements are described below. The return
When used inside a switch statement, break causes the switch statement to end its
statement is described in Functions, and the throw statement is described in
execution immediately and to transfer control to the code after the switch statement’s
Propagating Errors Using Throwing Functions.
closing brace (}).

Continue
This behavior can be used to match and ignore one or more cases in a switch
The continue statement tells a loop to stop what it is doing and start again at the statement. Because Swift’s switch statement is exhaustive and does not allow empty
beginning of the next iteration through the loop. It says “I am done with the current cases, it is sometimes necessary to deliberately match and ignore a case in order to
loop iteration” without leaving the loop altogether. make your intentions explicit. You do this by writing the break statement as the entire
body of the case you want to ignore. When that case is matched by the switch
The following example removes all vowels and spaces from a lowercase string to statement, the break statement inside the case ends the switch statement’s execution
create a cryptic puzzle phrase: immediately.

let puzzleInput = "great minds think alike" NOTE

var puzzleOutput = ""
A switch case that contains only a comment is reported as a compile-time error.
let charactersToRemove: [Character] = ["a", "e", "i", "o", "u", " Comments are not statements and do not cause a switch case to be ignored. Always
"] use a break statement to ignore a switch case.
for character in puzzleInput.characters {
The following example switches on a Character value and determines whether it
if charactersToRemove.contains(character) { represents a number symbol in one of four languages. For brevity, multiple values are
continue covered in a single switch case.
} else {
let numberSymbol: Character = "三" // Chinese symbol for the number
puzzleOutput.append(character)
3
}
var possibleIntegerValue: Int?
}
switch numberSymbol {
print(puzzleOutput)
case "1", "١", "⼀一", "๑":
// Prints "grtmndsthnklk"
possibleIntegerValue = 1
The code above calls the continue keyword whenever it matches a vowel or a space,
case "2", "٢", "⼆二", "๒":
causing the current iteration of the loop to end immediately and to jump straight to the
start of the next iteration. possibleIntegerValue = 2
case "3", "٣", "三", "๓":
Break possibleIntegerValue = 3
The break statement ends execution of an entire control flow statement immediately. case "4", "٤", "四", "๔":
The break statement can be used inside a switch statement or loop statement when
possibleIntegerValue = 4

56
 default: switch integerToDescribe {
break case 2, 3, 5, 7, 11, 13, 17, 19:
} description += " a prime number, and also"
if let integerValue = possibleIntegerValue { fallthrough
print("The integer value of \(numberSymbol) is \ default:
(integerValue).")
description += " an integer."
} else {
}
print("An integer value could not be found for \
print(description)
(numberSymbol).")
// Prints "The number 5 is a prime number, and also an integer."
}
// Prints "The integer value of 三 is 3." This example declares a new String variable called description and assigns it an
initial value. The function then considers the value of integerToDescribe using a
This example checks numberSymbol to determine whether it is a Latin, Arabic, switch statement. If the value of integerToDescribe is one of the prime numbers in
Chinese, or Thai symbol for the numbers 1 to 4. If a match is found, one of the switch the list, the function appends text to the end of description, to note that the number
statement’s cases sets an optional Int? variable called possibleIntegerValue to an is prime. It then uses the fallthrough keyword to “fall into” the default case as well.
appropriate integer value. The default case adds some extra text to the end of the description, and the switch
statement is complete.
After the switch statement completes its execution, the example uses optional binding
to determine whether a value was found. The possibleIntegerValue variable has an Unless the value of integerToDescribe is in the list of known prime numbers, it is not
implicit initial value of nil by virtue of being an optional type, and so the optional matched by the first switch case at all. Because there are no other specific cases,
binding will succeed only if possibleIntegerValue was set to an actual value by one integerToDescribe is matched by the default case.
of the switch statement’s first four cases.
After the switch statement has finished executing, the number’s description is printed
Because it’s not practical to list every possible Character value in the example above, using the print(_:separator:terminator:) function. In this example, the number 5 is
a default case handles any characters that are not matched. This default case does correctly identified as a prime number.
not need to perform any action, and so it is written with a single break statement as its
body. As soon as the default case is matched, the break statement ends the switch NOTE
statement’s execution, and code execution continues from the if let statement.
The fallthrough keyword does not check the case conditions for the switch case that
it causes execution to fall into. The fallthrough keyword simply causes code
Fallthrough execution to move directly to the statements inside the next case (or default case)
Switch statements in Swift don’t fall through the bottom of each case and into the next block, as in C’s standard switch statement behavior.
one. Instead, the entire switch statement completes its execution as soon as the first
matching case is completed. By contrast, C requires you to insert an explicit break Labeled Statements
statement at the end of every switch case to prevent fallthrough. Avoiding default In Swift, you can nest loops and conditional statements inside other loops and
fallthrough means that Swift switch statements are much more concise and conditional statements to create complex control flow structures. However, loops and
predictable than their counterparts in C, and thus they avoid executing multiple switch conditional statements can both use the break statement to end their execution
cases by mistake. prematurely. Therefore, it is sometimes useful to be explicit about which loop or
conditional statement you want a break statement to terminate. Similarly, if you have
If you need C-style fallthrough behavior, you can opt in to this behavior on a case-by- multiple nested loops, it can be useful to be explicit about which loop the continue
case basis with the fallthrough keyword. The example below uses fallthrough to statement should affect.
create a textual description of a number.
To achieve these aims, you can mark a loop statement or conditional statement with a
let integerToDescribe = 5 statement label. With a conditional statement, you can use a statement label with the
var description = "The number \(integerToDescribe) is" break statement to end the execution of the labeled statement. With a loop statement,

57
you can use a statement label with the break or continue statement to end or The while loop’s condition is while square != finalSquare, to reflect that you must
continue the execution of the labeled statement. land exactly on square 25.

A labeled statement is indicated by placing a label on the same line as the statement’s gameLoop: while square != finalSquare {
introducer keyword, followed by a colon. Here’s an example of this syntax for a while diceRoll += 1
loop, although the principle is the same for all loops and switch statements: if diceRoll == 7 { diceRoll = 1 }
switch square + diceRoll {
label name: while condition {
statements case finalSquare:
}
// diceRoll will move us to the final square, so the game
The following example uses the break and continue statements with a labeled while is over
loop for an adapted version of the Snakes and Ladders game that you saw earlier in break gameLoop
this chapter. This time around, the game has an extra rule:
case let newSquare where newSquare > finalSquare:
// diceRoll will move us beyond the final square, so roll
To win, you must land exactly on square 25. again
If a particular dice roll would take you beyond square 25, you must roll again until you continue gameLoop
roll the exact number needed to land on square 25. default:
// this is a valid move, so find out its effect
The game board is the same as before.
square += diceRoll
square += board[square]
}
}
print("Game over!")

The dice is rolled at the start of each loop. Rather than moving the player immediately,
the loop uses a switch statement to consider the result of the move and to determine
whether the move is allowed:

If the dice roll will move the player onto the final square, the game is over.
The break gameLoop statement transfers control to the first line of code
outside of the while loop, which ends the game.
The values of finalSquare, board, square, and diceRoll are initialized in the same
way as before: If the dice roll will move the player beyond the final square, the move is
invalid and the player needs to roll again. The continue gameLoop
let finalSquare = 25 statement ends the current while loop iteration and begins the next
var board = [Int](repeating: 0, count: finalSquare + 1) iteration of the loop.
board[03] = +08; board[06] = +11; board[09] = +09; board[10] = +02 In all other cases, the dice roll is a valid move. The player moves forward
board[14] = -10; board[19] = -11; board[22] = -02; board[24] = -08 by diceRoll squares, and the game logic checks for any snakes and
var square = 0
ladders. The loop then ends, and control returns to the while condition to
decide whether another turn is required.
var diceRoll = 0
NOTE
This version of the game uses a while loop and a switch statement to implement the
game’s logic. The while loop has a statement label called gameLoop to indicate that it If the break statement above did not use the gameLoop label, it would break out of the
switch statement, not the while statement. Using the gameLoop label makes it clear
is the main game loop for the Snakes and Ladders game.
which control statement should be terminated.

58
 It is not strictly necessary to use the gameLoop label when calling continue gameLoop If that condition is not met, the code inside the else branch is executed. That branch
to jump to the next iteration of the loop. There is only one loop in the game, and must transfer control to exit the code block in which the guard statement appears. It
therefore no ambiguity as to which loop the continue statement will affect. However,
can do this with a control transfer statement such as return, break, continue, or
there is no harm in using the gameLoop label with the continue statement. Doing so is
consistent with the label’s use alongside the break statement and helps make the throw, or it can call a function or method that doesn’t return, such as
game’s logic clearer to read and understand. fatalError(_:file:line:).

Early Exit Using a guard statement for requirements improves the readability of your code,
compared to doing the same check with an if statement. It lets you write the code
A guard statement, like an if statement, executes statements depending on the
that’s typically executed without wrapping it in an else block, and it lets you keep the
Boolean value of an expression. You use a guard statement to require that a condition
code that handles a violated requirement next to the requirement.
must be true in order for the code after the guard statement to be executed. Unlike an
if statement, a guard statement always has an else clause—the code inside the else
clause is executed if the condition is not true. Checking API Availability
Swift has built-in support for checking API availability, which ensures that you don’t
func greet(person: [String: String]) { accidentally use APIs that are unavailable on a given deployment target.
guard let name = person["name"] else {
return The compiler uses availability information in the SDK to verify that all of the APIs used
in your code are available on the deployment target specified by your project. Swift
}
reports an error at compile time if you try to use an API that isn’t available.

print("Hello \(name)!") You use an availability condition in an if or guard statement to conditionally execute a
block of code, depending on whether the APIs you want to use are available at
guard let location = person["location"] else { runtime. The compiler uses the information from the availability condition when it
verifies that the APIs in that block of code are available.
print("I hope the weather is nice near you.")
return if #available(iOS 10, macOS 10.12, *) {
} // Use iOS 10 APIs on iOS, and use macOS 10.12 APIs on macOS
} else {
print("I hope the weather is nice in \(location).") // Fall back to earlier iOS and macOS APIs
} }

The availability condition above specifies that on iOS, the body of the if executes only
greet(person: ["name": "John"]) on iOS 10 and later; on macOS, only on macOS 10.12 and later. The last argument, *,
// Prints "Hello John!" is required and specifies that on any other platform, the body of the if executes on
// Prints "I hope the weather is nice near you." the minimum deployment target specified by your target.
greet(person: ["name": "Jane", "location": "Cupertino"])
In its general form, the availability condition takes a list of platform names and
// Prints "Hello Jane!"
versions. You use platform names such as iOS, macOS, watchOS, and tvOS—for the full
// Prints "I hope the weather is nice in Cupertino." list, see Declaration Attributes. In addition to specifying major version numbers like
If the guard statement’s condition is met, code execution continues after the guard iOS 8, you can specify minor versions numbers like iOS 8.3 and macOS 10.10.3.
statement’s closing brace. Any variables or constants that were assigned values using
if #available(platform name version, ..., *) {
an optional binding as part of the condition are available for the rest of the code block
statements to execute if the APIs are available
that the guard statement appears in. } else {
fallback statements to execute if the APIs are unavailable
}

59
 return greeting
Section 6 }

Functions All of this information is rolled up into the function’s definition, which is prefixed with
the func keyword. You indicate the function’s return type with the return arrow -> (a
hyphen followed by a right angle bracket), which is followed by the name of the type
to return.

Functions The definition describes what the function does, what it expects to receive, and what it
returns when it is done. The definition makes it easy for the function to be called
Functions are self-contained chunks of code that perform a specific task. You give a unambiguously from elsewhere in your code:
function a name that identifies what it does, and this name is used to “call” the
function to perform its task when needed. print(greet(person: "Anna"))
// Prints "Hello, Anna!"
Swift’s unified function syntax is flexible enough to express anything from a simple C- print(greet(person: "Brian"))
style function with no parameter names to a complex Objective-C-style method with
// Prints "Hello, Brian!"
names and argument labels for each parameter. Parameters can provide default
values to simplify function calls and can be passed as in-out parameters, which You call the greet(person:) function by passing it a String value after the person
modify a passed variable once the function has completed its execution. argument label, such as greet(person: "Anna"). Because the function returns a
String value, greet(person:) can be wrapped in a call to the
Every function in Swift has a type, consisting of the function’s parameter types and print(_:separator:terminator:) function to print that string and see its return value,
return type. You can use this type like any other type in Swift, which makes it easy to as shown above.
pass functions as parameters to other functions, and to return functions from
functions. Functions can also be written within other functions to encapsulate useful NOTE
functionality within a nested function scope. The print(_:separator:terminator:) function doesn’t have a label for its first
argument, and its other arguments are optional because they have a default value.
Defining and Calling Functions These variations on function syntax are discussed below in Function Argument Labels
and Parameter Names and Default Parameter Values.
When you define a function, you can optionally define one or more named, typed
values that the function takes as input, known as parameters. You can also optionally The body of the greet(person:) function starts by defining a new String constant
define a type of value that the function will pass back as output when it is done, called greeting and setting it to a simple greeting message. This greeting is then
known as its return type. passed back out of the function using the return keyword. In the line of code that
says return greeting, the function finishes its execution and returns the current
Every function has a function name, which describes the task that the function value of greeting.
performs. To use a function, you “call” that function with its name and pass it input
values (known as arguments) that match the types of the function’s parameters. A You can call the greet(person:) function multiple times with different input values.
function’s arguments must always be provided in the same order as the function’s The example above shows what happens if it is called with an input value of "Anna",
parameter list. and an input value of "Brian". The function returns a tailored greeting in each case.

The function in the example below is called greet(person:), because that’s what it To make the body of this function shorter, you can combine the message creation and
does—it takes a person’s name as input and returns a greeting for that person. To the return statement into one line:
accomplish this, you define one input parameter—a String value called person—and
a return type of String, which will contain a greeting for that person: func greetAgain(person: String) -> String {
return "Hello again, " + person + "!"
func greet(person: String) -> String { }
let greeting = "Hello, " + person + "!" print(greetAgain(person: "Anna"))

60
 // Prints "Hello again, Anna!" Functions Without Return Values
Functions are not required to define a return type. Here’s a version of the
Function Parameters and Return Values greet(person:) function, which prints its own String value rather than returning it:
Function parameters and return values are extremely flexible in Swift. You can define
anything from a simple utility function with a single unnamed parameter to a complex func greet(person: String) {
function with expressive parameter names and different parameter options. print("Hello, \(person)!")
}
Functions Without Parameters greet(person: "Dave")
Functions are not required to define input parameters. Here’s a function with no input
// Prints "Hello, Dave!"
parameters, which always returns the same String message whenever it is called:
Because it does not need to return a value, the function’s definition does not include
func sayHelloWorld() -> String { the return arrow (->) or a return type.
return "hello, world"
NOTE
}
Strictly speaking, this version of the greet(person:) function does still return a value,
print(sayHelloWorld())
even though no return value is defined. Functions without a defined return type return a
// Prints "hello, world" special value of type Void. This is simply an empty tuple, which is written as ().
The function definition still needs parentheses after the function’s name, even though The return value of a function can be ignored when it is called:
it does not take any parameters. The function name is also followed by an empty pair
of parentheses when the function is called. func printAndCount(string: String) -> Int {
print(string)
Functions With Multiple Parameters
return string.characters.count
Functions can have multiple input parameters, which are written within the function’s
}
parentheses, separated by commas.
func printWithoutCounting(string: String) {

This function takes a person’s name and whether they have already been greeted as let _ = printAndCount(string: string)
input, and returns an appropriate greeting for that person: }
printAndCount(string: "hello, world")
func greet(person: String, alreadyGreeted: Bool) -> String {
// prints "hello, world" and returns a value of 12
if alreadyGreeted {
printWithoutCounting(string: "hello, world")
return greetAgain(person: person)
// prints "hello, world" but does not return a value
} else {
The first function, printAndCount(string:), prints a string, and then returns its
return greet(person: person)
character count as an Int. The second function, printWithoutCounting(string:),
}
calls the first function, but ignores its return value. When the second function is called,
} the message is still printed by the first function, but the returned value is not used.
print(greet(person: "Tim", alreadyGreeted: true))
NOTE
// Prints "Hello again, Tim!"
Return values can be ignored, but a function that says it will return a value must always
You call the greet(person:alreadyGreeted:) function by passing it both a String
do so. A function with a defined return type cannot allow control to fall out of the bottom
argument value labeled person and a Bool argument value labeled alreadyGreeted in of the function without returning a value, and attempting to do so will result in a
parentheses, separated by commas. Note that this function is distinct from the compile-time error.
greet(person:) function shown in an earlier section. Although both functions have
names that begin with greet, the greet(person:alreadyGreeted:) function takes two Functions with Multiple Return Values
arguments but the greet(person:) function takes only one.

61
You can use a tuple type as the return type for a function to return multiple values as If the tuple type to be returned from a function has the potential to have “no value” for
part of one compound return value. the entire tuple, you can use an optional tuple return type to reflect the fact that the
entire tuple can be nil. You write an optional tuple return type by placing a question
The example below defines a function called minMax(array:), which finds the mark after the tuple type’s closing parenthesis, such as (Int, Int)? or (String,
smallest and largest numbers in an array of Int values: Int, Bool)?.

func minMax(array: [Int]) -> (min: Int, max: Int) { NOTE

var currentMin = array[0] An optional tuple type such as (Int, Int)? is different from a tuple that contains
var currentMax = array[0] optional types such as (Int?, Int?). With an optional tuple type, the entire tuple is
optional, not just each individual value within the tuple.
for value in array[1..<array.count] {
if value < currentMin { The minMax(array:) function above returns a tuple containing two Int values.
However, the function does not perform any safety checks on the array it is passed. If
currentMin = value
the array argument contains an empty array, the minMax(array:) function, as defined
} else if value > currentMax { above, will trigger a runtime error when attempting to access array[0].
currentMax = value
} To handle an empty array safely, write the minMax(array:) function with an optional
} tuple return type and return a value of nil when the array is empty:
return (currentMin, currentMax)
func minMax(array: [Int]) -> (min: Int, max: Int)? {
}
if array.isEmpty { return nil }
The minMax(array:) function returns a tuple containing two Int values. These values var currentMin = array[0]
are labeled min and max so that they can be accessed by name when querying the
var currentMax = array[0]
function’s return value.
for value in array[1..<array.count] {

The body of the minMax(array:) function starts by setting two working variables called if value < currentMin {
currentMin and currentMax to the value of the first integer in the array. The function currentMin = value
then iterates over the remaining values in the array and checks each value to see if it } else if value > currentMax {
is smaller or larger than the values of currentMin and currentMax respectively.
currentMax = value
Finally, the overall minimum and maximum values are returned as a tuple of two Int
values. }
}
Because the tuple’s member values are named as part of the function’s return type, return (currentMin, currentMax)
they can be accessed with dot syntax to retrieve the minimum and maximum found }
values:
You can use optional binding to check whether this version of the minMax(array:)
let bounds = minMax(array: [8, -6, 2, 109, 3, 71]) function returns an actual tuple value or nil:
print("min is \(bounds.min) and max is \(bounds.max)")
if let bounds = minMax(array: [8, -6, 2, 109, 3, 71]) {
// Prints "min is -6 and max is 109"
print("min is \(bounds.min) and max is \(bounds.max)")
Note that the tuple’s members do not need to be named at the point that the tuple is }
returned from the function, because their names are already specified as part of the
// Prints "min is -6 and max is 109"
function’s return type.

Optional Tuple Return Types Function Argument Labels and Parameter Names

62
Each function parameter has both an argument label and a parameter name. The func someFunction(_ firstParameterName: Int, secondParameterName:
Int) {
argument label is used when calling the function; each argument is written in the
function call with its argument label before it. The parameter name is used in the // In the function body, firstParameterName and
secondParameterName
implementation of the function. By default, parameters use their parameter name as
their argument label. // refer to the argument values for the first and second
parameters.
func someFunction(firstParameterName: Int, secondParameterName: }
Int) {
someFunction(1, secondParameterName: 2)
// In the function body, firstParameterName and
secondParameterName If a parameter has an argument label, the argument must be labeled when you call the
function.
// refer to the argument values for the first and second
parameters.
}
Default Parameter Values
someFunction(firstParameterName: 1, secondParameterName: 2)
You can define a default value for any parameter in a function by assigning a value to
the parameter after that parameter’s type. If a default value is defined, you can omit
All parameters must have unique names. Although it’s possible for multiple that parameter when calling the function.
parameters to have the same argument label, unique argument labels help make your
code more readable. func someFunction(parameterWithoutDefault: Int,
parameterWithDefault: Int = 12) {
Specifying Argument Labels // If you omit the second argument when calling this function,
then
You write an argument label before the parameter name, separated by a space:
// the value of parameterWithDefault is 12 inside the function
func someFunction(argumentLabel parameterName: Int) { body.

// In the function body, parameterName refers to the argument }

value someFunction(parameterWithoutDefault: 3, parameterWithDefault:
// for that parameter. 6) // parameterWithDefault is 6

} someFunction(parameterWithoutDefault: 4) // parameterWithDefault is
12
Here’s a variation of the greet(person:) function that takes a person’s name and
Place parameters that don’t have default values at the beginning of a function’s
hometown and returns a greeting:
parameter list, before the parameters that have default values. Parameters that don’t
func greet(person: String, from hometown: String) -> String {
have default values are usually more important to the function’s meaning—writing
them first makes it easier to recognize that the same function is being called,
return "Hello \(person)! Glad you could visit from \
(hometown)."
regardless of whether any default parameters are omitted.
}
Variadic Parameters
print(greet(person: "Bill", from: "Cupertino"))
A variadic parameter accepts zero or more values of a specified type. You use a
// Prints "Hello Bill! Glad you could visit from Cupertino." variadic parameter to specify that the parameter can be passed a varying number of
The use of argument labels can allow a function to be called in an expressive, input values when the function is called. Write variadic parameters by inserting three
sentence-like manner, while still providing a function body that is readable and clear in period characters (...) after the parameter’s type name.
intent.
The values passed to a variadic parameter are made available within the function’s
Omitting Argument Labels body as an array of the appropriate type. For example, a variadic parameter with a
If you don’t want an argument label for a parameter, write an underscore (_) instead of name of numbers and a type of Double... is made available within the function’s body
an explicit argument label for that parameter. as a constant array called numbers of type [Double].

63
The example below calculates the arithmetic mean (also known as the average) for a func swapTwoInts(_ a: inout Int, _ b: inout Int) {
list of numbers of any length: let temporaryA = a
a = b
func arithmeticMean(_ numbers: Double...) -> Double {
b = temporaryA
var total: Double = 0
}
for number in numbers {
The swapTwoInts(_:_:) function simply swaps the value of b into a, and the value of a
total += number
into b. The function performs this swap by storing the value of a in a temporary
} constant called temporaryA, assigning the value of b to a, and then assigning
return total / Double(numbers.count) temporaryA to b.
}
arithmeticMean(1, 2, 3, 4, 5) You can call the swapTwoInts(_:_:) function with two variables of type Int to swap
their values. Note that the names of someInt and anotherInt are prefixed with an
// returns 3.0, which is the arithmetic mean of these five numbers
ampersand when they are passed to the swapTwoInts(_:_:) function:
arithmeticMean(3, 8.25, 18.75)
// returns 10.0, which is the arithmetic mean of these three numbers var someInt = 3
NOTE var anotherInt = 107

A function may have at most one variadic parameter. swapTwoInts(&someInt, &anotherInt)

print("someInt is now \(someInt), and anotherInt is now \
In-Out Parameters (anotherInt)")
Function parameters are constants by default. Trying to change the value of a function // Prints "someInt is now 107, and anotherInt is now 3"
parameter from within the body of that function results in a compile-time error. This The example above shows that the original values of someInt and anotherInt are
means that you can’t change the value of a parameter by mistake. If you want a modified by the swapTwoInts(_:_:) function, even though they were originally defined
function to modify a parameter’s value, and you want those changes to persist after outside of the function.
the function call has ended, define that parameter as an in-out parameter instead.
NOTE
You write an in-out parameter by placing the inout keyword right before a parameter’s
In-out parameters are not the same as returning a value from a function. The
type. An in-out parameter has a value that is passed in to the function, is modified by
swapTwoInts example above does not define a return type or return a value, but it still
the function, and is passed back out of the function to replace the original value. For a modifies the values of someInt and anotherInt. In-out parameters are an alternative
detailed discussion of the behavior of in-out parameters and associated compiler way for a function to have an effect outside of the scope of its function body.
optimizations, see In-Out Parameters.
Function Types
You can only pass a variable as the argument for an in-out parameter. You cannot
Every function has a specific function type, made up of the parameter types and the
pass a constant or a literal value as the argument, because constants and literals
return type of the function.
cannot be modified. You place an ampersand (&) directly before a variable’s name
when you pass it as an argument to an in-out parameter, to indicate that it can be
For example:
modified by the function.
func addTwoInts(_ a: Int, _ b: Int) -> Int {
NOTE
return a + b
In-out parameters cannot have default values, and variadic parameters cannot be
marked as inout. }
func multiplyTwoInts(_ a: Int, _ b: Int) -> Int {
Here’s an example of a function called swapTwoInts(_:_:), which has two in-out
return a * b
integer parameters called a and b:
}

64
This example defines two simple mathematical functions called addTwoInts and As with any other type, you can leave it to Swift to infer the function type when you
multiplyTwoInts. These functions each take two Int values, and return an Int value, assign a function to a constant or variable:
which is the result of performing an appropriate mathematical operation.
let anotherMathFunction = addTwoInts
The type of both of these functions is (Int, Int) -> Int. This can be read as: // anotherMathFunction is inferred to be of type (Int, Int) -> Int

“A function type that has two parameters, both of type Int, and that returns a value of Function Types as Parameter Types
type Int.” You can use a function type such as (Int, Int) -> Int as a parameter type for
another function. This enables you to leave some aspects of a function’s
Here’s another example, for a function with no parameters or return value: implementation for the function’s caller to provide when the function is called.

func printHelloWorld() { Here’s an example to print the results of the math functions from above:
print("hello, world")
func printMathResult(_ mathFunction: (Int, Int) -> Int, _ a: Int, _
}
b: Int) {
The type of this function is () -> Void, or “a function that has no parameters, and print("Result: \(mathFunction(a, b))")
returns Void.”
}
printMathResult(addTwoInts, 3, 5)
Using Function Types
// Prints "Result: 8"
You use function types just like any other types in Swift. For example, you can define
a constant or variable to be of a function type and assign an appropriate function to This example defines a function called printMathResult(_:_:_:), which has three
that variable: parameters. The first parameter is called mathFunction, and is of type (Int, Int) ->
Int. You can pass any function of that type as the argument for this first parameter.
var mathFunction: (Int, Int) -> Int = addTwoInts The second and third parameters are called a and b, and are both of type Int. These
This can be read as: are used as the two input values for the provided math function.

“Define a variable called mathFunction, which has a type of ‘a function that takes two When printMathResult(_:_:_:) is called, it is passed the addTwoInts(_:_:)
Int values, and returns an Int value.’ Set this new variable to refer to the function function, and the integer values 3 and 5. It calls the provided function with the values 3
called addTwoInts.” and 5, and prints the result of 8.

The addTwoInts(_:_:) function has the same type as the mathFunction variable, and The role of printMathResult(_:_:_:) is to print the result of a call to a math function
so this assignment is allowed by Swift’s type-checker. of an appropriate type. It doesn’t matter what that function’s implementation actually
does—it matters only that the function is of the correct type. This enables
printMathResult(_:_:_:) to hand off some of its functionality to the caller of the
You can now call the assigned function with the name mathFunction:
function in a type-safe way.
print("Result: \(mathFunction(2, 3))")
Function Types as Return Types
// Prints "Result: 5"
You can use a function type as the return type of another function. You do this by
A different function with the same matching type can be assigned to the same writing a complete function type immediately after the return arrow (->) of the returning
variable, in the same way as for non-function types: function.
mathFunction = multiplyTwoInts
The next example defines two simple functions called stepForward(_:) and
print("Result: \(mathFunction(2, 3))")
stepBackward(_:). The stepForward(_:) function returns a value one more than its
// Prints "Result: 6" input value, and the stepBackward(_:) function returns a value one less than its input
value. Both functions have a type of (Int) -> Int:

65
 func stepForward(_ input: Int) -> Int {
Nested Functions
return input + 1
All of the functions you have encountered so far in this chapter have been examples
} of global functions, which are defined at a global scope. You can also define functions
func stepBackward(_ input: Int) -> Int { inside the bodies of other functions, known as nested functions.
return input - 1
}
Nested functions are hidden from the outside world by default, but can still be called
and used by their enclosing function. An enclosing function can also return one of its
Here’s a function called chooseStepFunction(backward:), whose return type is (Int) nested functions to allow the nested function to be used in another scope.
-> Int. The chooseStepFunction(backward:) function returns the stepForward(_:)
function or the stepBackward(_:) function based on a Boolean parameter called You can rewrite the chooseStepFunction(backward:) example above to use and
backward: return nested functions:
func chooseStepFunction(backward: Bool) -> (Int) -> Int { func chooseStepFunction(backward: Bool) -> (Int) -> Int {
return backward ? stepBackward : stepForward func stepForward(input: Int) -> Int { return input + 1 }
} func stepBackward(input: Int) -> Int { return input - 1 }
You can now use chooseStepFunction(backward:) to obtain a function that will step return backward ? stepBackward : stepForward
in one direction or the other: }
var currentValue = -4
var currentValue = 3
let moveNearerToZero = chooseStepFunction(backward: currentValue >
let moveNearerToZero = chooseStepFunction(backward: currentValue >
0)
0)
// moveNearerToZero now refers to the nested stepForward() function
// moveNearerToZero now refers to the stepBackward() function
while currentValue != 0 {
The preceding example determines whether a positive or negative step is needed to
move a variable called currentValue progressively closer to zero. currentValue has print("\(currentValue)... ")
an initial value of 3, which means that currentValue > 0 returns true, causing currentValue = moveNearerToZero(currentValue)
chooseStepFunction(backward:) to return the stepBackward(_:) function. A }
reference to the returned function is stored in a constant called moveNearerToZero. print("zero!")
// -4...
Now that moveNearerToZero refers to the correct function, it can be used to count to
zero: // -3...
// -2...
print("Counting to zero:") // -1...
// Counting to zero: // zero!
while currentValue != 0 {
print("\(currentValue)... ")
currentValue = moveNearerToZero(currentValue)
}
print("zero!")
// 3...
// 2...
// 1...
// zero!

66
 Section 7 However, it is sometimes useful to write shorter versions of function-like constructs
without a full declaration and name. This is particularly true when you work with

Closures functions or methods that take functions as one or more of their arguments.

Closure expressions are a way to write inline closures in a brief, focused syntax.
Closure expressions provide several syntax optimizations for writing closures in a
shortened form without loss of clarity or intent. The closure expression examples
Closures below illustrate these optimizations by refining a single example of the sorted(by:)
method over several iterations, each of which expresses the same functionality in a
Closures are self-contained blocks of functionality that can be passed around and more succinct way.
used in your code. Closures in Swift are similar to blocks in C and Objective-C and to
lambdas in other programming languages. The Sorted Method
Swift’s standard library provides a method called sorted(by:), which sorts an array of
Closures can capture and store references to any constants and variables from the values of a known type, based on the output of a sorting closure that you provide.
context in which they are defined. This is known as closing over those constants and Once it completes the sorting process, the sorted(by:) method returns a new array
variables. Swift handles all of the memory management of capturing for you. of the same type and size as the old one, with its elements in the correct sorted order.
The original array is not modified by the sorted(by:) method.
NOTE

Don’t worry if you are not familiar with the concept of capturing. It is explained in detail The closure expression examples below use the sorted(by:) method to sort an array
below in Capturing Values. of String values in reverse alphabetical order. Here’s the initial array to be sorted:

Global and nested functions, as introduced in Functions, are actually special cases of let names = ["Chris", "Alex", "Ewa", "Barry", "Daniella"]
closures. Closures take one of three forms:
The sorted(by:) method accepts a closure that takes two arguments of the same
type as the array’s contents, and returns a Bool value to say whether the first value
Global functions are closures that have a name and do not capture any
should appear before or after the second value once the values are sorted. The
values.
sorting closure needs to return true if the first value should appear before the second
Nested functions are closures that have a name and can capture values value, and false otherwise.
from their enclosing function.
Closure expressions are unnamed closures written in a lightweight syntax This example is sorting an array of String values, and so the sorting closure needs to
that can capture values from their surrounding context. be a function of type (String, String) -> Bool.

Swift’s closure expressions have a clean, clear style, with optimizations that One way to provide the sorting closure is to write a normal function of the correct type,
encourage brief, clutter-free syntax in common scenarios. These optimizations and to pass it in as an argument to the sorted(by:) method:
include:
func backward(_ s1: String, _ s2: String) -> Bool {
Inferring parameter and return value types from context return s1 > s2
Implicit returns from single-expression closures }
Shorthand argument names var reversedNames = names.sorted(by: backward)

Trailing closure syntax // reversedNames is equal to ["Ewa", "Daniella", "Chris", "Barry",

"Alex"]

Closure Expressions If the first string (s1) is greater than the second string (s2), the backward(_:_:)
function will return true, indicating that s1 should appear before s2 in the sorted array.
Nested functions, as introduced in Nested Functions, are a convenient means of
For characters in strings, “greater than” means “appears later in the alphabet than”.
naming and defining self-contained blocks of code as part of a larger function.
This means that the letter "B" is “greater than” the letter "A", and the string "Tom" is

67
greater than the string "Tim". This gives a reverse alphabetical sort, with "Barry" is being called on an array of strings, so its argument must be a function of type
being placed before "Alex", and so on. (String, String) -> Bool. This means that the (String, String) and Bool types
do not need to be written as part of the closure expression’s definition. Because all of
However, this is a rather long-winded way to write what is essentially a single- the types can be inferred, the return arrow (->) and the parentheses around the
expression function (a > b). In this example, it would be preferable to write the sorting names of the parameters can also be omitted:
closure inline, using closure expression syntax.
reversedNames = names.sorted(by: { s1, s2 in return s1 > s2 } )
Closure Expression Syntax It is always possible to infer the parameter types and return type when passing a
Closure expression syntax has the following general form: closure to a function or method as an inline closure expression. As a result, you never
need to write an inline closure in its fullest form when the closure is used as a function
{ (parameters) -> return type in or method argument.
statements
}
Nonetheless, you can still make the types explicit if you wish, and doing so is
The parameters in closure expression syntax can be in-out parameters, but they can’t encouraged if it avoids ambiguity for readers of your code. In the case of the
have a default value. Variadic parameters can be used if you name the variadic sorted(by:) method, the purpose of the closure is clear from the fact that sorting is
parameter. Tuples can also be used as parameter types and return types. taking place, and it is safe for a reader to assume that the closure is likely to be
working with String values, because it is assisting with the sorting of an array of
The example below shows a closure expression version of the backward(_:_:) strings.
function from earlier:
Implicit Returns from Single-Expression Closures
reversedNames = names.sorted(by: { (s1: String, s2: String) -> Bool
in Single-expression closures can implicitly return the result of their single expression by
return s1 > s2 omitting the return keyword from their declaration, as in this version of the previous
example:
})

Note that the declaration of parameters and return type for this inline closure is reversedNames = names.sorted(by: { s1, s2 in s1 > s2 } )
identical to the declaration from the backward(_:_:) function. In both cases, it is Here, the function type of the sorted(by:) method’s argument makes it clear that a
written as (s1: String, s2: String) -> Bool. However, for the inline closure Bool value must be returned by the closure. Because the closure’s body contains a
expression, the parameters and return type are written inside the curly braces, not single expression (s1 > s2) that returns a Bool value, there is no ambiguity, and the
outside of them. return keyword can be omitted.

The start of the closure’s body is introduced by the in keyword. This keyword Shorthand Argument Names
indicates that the definition of the closure’s parameters and return type has finished,
Swift automatically provides shorthand argument names to inline closures, which can
and the body of the closure is about to begin.
be used to refer to the values of the closure’s arguments by the names $0, $1, $2, and
so on.
Because the body of the closure is so short, it can even be written on a single line:

reversedNames = names.sorted(by: { (s1: String, s2: String) -> Bool If you use these shorthand argument names within your closure expression, you can
in return s1 > s2 } ) omit the closure’s argument list from its definition, and the number and type of the
shorthand argument names will be inferred from the expected function type. The in
This illustrates that the overall call to the sorted(by:) method has remained the
keyword can also be omitted, because the closure expression is made up entirely of
same. A pair of parentheses still wrap the entire argument for the method. However,
its body:
that argument is now an inline closure.
reversedNames = names.sorted(by: { $0 > $1 } )
Inferring Type From Context
Here, $0 and $1 refer to the closure’s first and second String arguments.
Because the sorting closure is passed as an argument to a method, Swift can infer the
types of its parameters and the type of the value it returns. The sorted(by:) method

68
Operator Methods If a closure expression is provided as the function or method’s only argument and you
There’s actually an even shorter way to write the closure expression above. Swift’s provide that expression as a trailing closure, you do not need to write a pair of
String type defines its string-specific implementation of the greater-than operator (>) parentheses () after the function or method’s name when you call the function:
as a method that has two parameters of type String, and returns a value of type Bool.
This exactly matches the method type needed by the sorted(by:) method. Therefore, reversedNames = names.sorted { $0 > $1 }
you can simply pass in the greater-than operator, and Swift will infer that you want to Trailing closures are most useful when the closure is sufficiently long that it is not
use its string-specific implementation: possible to write it inline on a single line. As an example, Swift’s Array type has a
map(_:) method which takes a closure expression as its single argument. The closure
reversedNames = names.sorted(by: >) is called once for each item in the array, and returns an alternative mapped value
For more about operator method, see Operator Methods. (possibly of some other type) for that item. The nature of the mapping and the type of
the returned value is left up to the closure to specify.
Trailing Closures
After applying the provided closure to each array element, the map(_:) method returns
If you need to pass a closure expression to a function as the function’s final argument a new array containing all of the new mapped values, in the same order as their
and the closure expression is long, it can be useful to write it as a trailing closure corresponding values in the original array.
instead. A trailing closure is written after the function call’s parentheses, even though it
is still an argument to the function. When you use the trailing closure syntax, you don’t
Here’s how you can use the map(_:) method with a trailing closure to convert an array
write the argument label for the closure as part of the function call.
of Int values into an array of String values. The array [16, 58, 510] is used to
create the new array ["OneSix", "FiveEight", "FiveOneZero"]:
func someFunctionThatTakesAClosure(closure: () -> Void) {
// function body goes here let digitNames = [
} 0: "Zero", 1: "One", 2: "Two", 3: "Three", 4: "Four",
5: "Five", 6: "Six", 7: "Seven", 8: "Eight", 9: "Nine"
// Here's how you call this function without using a trailing ]
closure:
let numbers = [16, 58, 510]

someFunctionThatTakesAClosure(closure: {
The code above creates a dictionary of mappings between the integer digits and
English-language versions of their names. It also defines an array of integers, ready to
// closure's body goes here
be converted into strings.
})
You can now use the numbers array to create an array of String values, by passing a
// Here's how you call this function with a trailing closure closure expression to the array’s map(_:) method as a trailing closure:
instead:
let strings = numbers.map {

someFunctionThatTakesAClosure() { (number) -> String in

// trailing closure's body goes here var number = number

} var output = ""

repeat {
The string-sorting closure from the Closure Expression Syntax section above can be
written outside of the sorted(by:) method’s parentheses as a trailing closure: output = digitNames[number % 10]! + output
number /= 10
reversedNames = names.sorted() { $0 > $1 } } while number > 0
return output
}

69
 // strings is inferred to be of type [String] In Swift, the simplest form of a closure that can capture values is a nested function,
// its value is ["OneSix", "FiveEight", "FiveOneZero"] written within the body of another function. A nested function can capture any of its
outer function’s arguments and can also capture any constants and variables defined
The map(_:) method calls the closure expression once for each item in the array. You
within the outer function.
do not need to specify the type of the closure’s input parameter, number, because the
type can be inferred from the values in the array to be mapped.
Here’s an example of a function called makeIncrementer, which contains a nested
function called incrementer. The nested incrementer() function captures two values,
In this example, the variable number is initialized with the value of the closure’s number
runningTotal and amount, from its surrounding context. After capturing these values,
parameter, so that the value can be modified within the closure body. (The parameters
incrementer is returned by makeIncrementer as a closure that increments
to functions and closures are always constants.) The closure expression also specifies
runningTotal by amount each time it is called.
a return type of String, to indicate the type that will be stored in the mapped output
array.
func makeIncrementer(forIncrement amount: Int) -> () -> Int {
var runningTotal = 0
The closure expression builds a string called output each time it is called. It calculates
the last digit of number by using the remainder operator (number % 10), and uses this func incrementer() -> Int {
digit to look up an appropriate string in the digitNames dictionary. The closure can be runningTotal += amount
used to create a string representation of any integer greater than zero. return runningTotal
}
NOTE
return incrementer
The call to the digitNames dictionary’s subscript is followed by an exclamation mark
(!), because dictionary subscripts return an optional value to indicate that the dictionary }
lookup can fail if the key does not exist. In the example above, it is guaranteed that The return type of makeIncrementer is () -> Int. This means that it returns a
number % 10 will always be a valid subscript key for the digitNames dictionary, and so
function, rather than a simple value. The function it returns has no parameters, and
an exclamation mark is used to force-unwrap the String value stored in the subscript’s
optional return value. returns an Int value each time it is called. To learn how functions can return other
functions, see Function Types as Return Types.
The string retrieved from the digitNames dictionary is added to the front of output,
effectively building a string version of the number in reverse. (The expression number The makeIncrementer(forIncrement:) function defines an integer variable called
% 10 gives a value of 6 for 16, 8 for 58, and 0 for 510.) runningTotal, to store the current running total of the incrementer that will be
returned. This variable is initialized with a value of 0.
The number variable is then divided by 10. Because it is an integer, it is rounded down
during the division, so 16 becomes 1, 58 becomes 5, and 510 becomes 51. The makeIncrementer(forIncrement:) function has a single Int parameter with an
argument label of forIncrement, and a parameter name of amount. The argument
The process is repeated until number is equal to 0, at which point the output string is value passed to this parameter specifies how much runningTotal should be
returned by the closure, and is added to the output array by the map(_:) method. incremented by each time the returned incrementer function is called. The
makeIncrementer function defines a nested function called incrementer, which
The use of trailing closure syntax in the example above neatly encapsulates the performs the actual incrementing. This function simply adds amount to runningTotal,
closure’s functionality immediately after the function that closure supports, without and returns the result.
needing to wrap the entire closure within the map(_:) method’s outer parentheses.
When considered in isolation, the nested incrementer() function might seem
Capturing Values unusual:
A closure can capture constants and variables from the surrounding context in which it
func incrementer() -> Int {
is defined. The closure can then refer to and modify the values of those constants and
variables from within its body, even if the original scope that defined the constants and runningTotal += amount
variables no longer exists. return runningTotal
}

70
The incrementer() function doesn’t have any parameters, and yet it refers to strong reference cycles. For more information, see Strong Reference Cycles for
runningTotal and amount from within its function body. It does this by capturing a Closures.
reference to runningTotal and amount from the surrounding function and using them
within its own function body. Capturing by reference ensures that runningTotal and Closures Are Reference Types
amount do not disappear when the call to makeIncrementer ends, and also ensures In the example above, incrementBySeven and incrementByTen are constants, but the
that runningTotal is available the next time the incrementer function is called. closures these constants refer to are still able to increment the runningTotal variables
that they have captured. This is because functions and closures are reference types.
NOTE

As an optimization, Swift may instead capture and store a copy of a value if that value Whenever you assign a function or a closure to a constant or a variable, you are
is not mutated by a closure, and if the value is not mutated after the closure is created. actually setting that constant or variable to be a reference to the function or closure. In
Swift also handles all memory management involved in disposing of variables when the example above, it is the choice of closure that incrementByTen refers to that is
they are no longer needed. constant, and not the contents of the closure itself.

Here’s an example of makeIncrementer in action: This also means that if you assign a closure to two different constants or variables,
both of those constants or variables will refer to the same closure:
let incrementByTen = makeIncrementer(forIncrement: 10)

This example sets a constant called incrementByTen to refer to an incrementer let alsoIncrementByTen = incrementByTen
function that adds 10 to its runningTotal variable each time it is called. Calling the alsoIncrementByTen()
function multiple times shows this behavior in action: // returns a value of 50

incrementByTen()
Escaping Closures
// returns a value of 10
A closure is said to escape a function when the closure is passed as an argument to
incrementByTen()
the function, but is called after the function returns. When you declare a function that
// returns a value of 20 takes a closure as one of its parameters, you can write @escaping before the
incrementByTen() parameter’s type to indicate that the closure is allowed to escape.
// returns a value of 30
One way that a closure can escape is by being stored in a variable that is defined
If you create a second incrementer, it will have its own stored reference to a new,
outside the function. As an example, many functions that start an asynchronous
separate runningTotal variable:
operation take a closure argument as a completion handler. The function returns after
let incrementBySeven = makeIncrementer(forIncrement: 7)
it starts the operation, but the closure isn’t called until the operation is completed—the
closure needs to escape, to be called later. For example:
incrementBySeven()
// returns a value of 7 var completionHandlers: [() -> Void] = []
Calling the original incrementer (incrementByTen) again continues to increment its func someFunctionWithEscapingClosure(completionHandler: @escaping
own runningTotal variable, and does not affect the variable captured by () -> Void) {
incrementBySeven: completionHandlers.append(completionHandler)
}
incrementByTen()
The someFunctionWithEscapingClosure(_:) function takes a closure as its argument
// returns a value of 40
and adds it to an array that’s declared outside the function. If you didn’t mark the
NOTE parameter of this function with @escaping, you would get a compiler error.
If you assign a closure to a property of a class instance, and the closure captures that
instance by referring to the instance or its members, you will create a strong reference Marking a closure with @escaping means you have to refer to self explicitly within the
cycle between the closure and the instance. Swift uses capture lists to break these closure. For example, in the code below, the closure passed to
someFunctionWithEscapingClosure(_:) is an escaping closure, which means it

71
needs to refer to self explicitly. In contrast, the closure passed to computationally expensive, because it lets you control when that code is evaluated.
someFunctionWithNonescapingClosure(_:) is a nonescaping closure, which means it The code below shows how a closure delays evaluation.
can refer to self implicitly.
var customersInLine = ["Chris", "Alex", "Ewa", "Barry", "Daniella"]
func someFunctionWithNonescapingClosure(closure: () -> Void) { print(customersInLine.count)
closure() // Prints "5"
}

let customerProvider = { customersInLine.remove(at: 0) }

class SomeClass { print(customersInLine.count)
var x = 10 // Prints "5"
func doSomething() {
someFunctionWithEscapingClosure { self.x = 100 } print("Now serving \(customerProvider())!")
someFunctionWithNonescapingClosure { x = 200 } // Prints "Now serving Chris!"
} print(customersInLine.count)
} // Prints "4"

Even though the first element of the customersInLine array is removed by the code
let instance = SomeClass() inside the closure, the array element isn’t removed until the closure is actually called.
instance.doSomething() If the closure is never called, the expression inside the closure is never evaluated,
print(instance.x) which means the array element is never removed. Note that the type of
customerProvider is not String but () -> String—a function with no parameters
// Prints "200"
that returns a string.

completionHandlers.first?() You get the same behavior of delayed evaluation when you pass a closure as an
print(instance.x) argument to a function.
// Prints "100"
// customersInLine is ["Alex", "Ewa", "Barry", "Daniella"]

Autoclosures func serve(customer customerProvider: () -> String) {

print("Now serving \(customerProvider())!")
An autoclosure is a closure that is automatically created to wrap an expression that’s
being passed as an argument to a function. It doesn’t take any arguments, and when }
it’s called, it returns the value of the expression that’s wrapped inside of it. This serve(customer: { customersInLine.remove(at: 0) } )
syntactic convenience lets you omit braces around a function’s parameter by writing a // Prints "Now serving Alex!"
normal expression instead of an explicit closure.
The serve(customer:) function in the listing above takes an explicit closure that
It’s common to call functions that take autoclosures, but it’s not common to implement returns a customer’s name. The version of serve(customer:) below performs the
that kind of function. For example, the assert(condition:message:file:line:) same operation but, instead of taking an explicit closure, it takes an autoclosure by
function takes an autoclosure for its condition and message parameters; its condition marking its parameter’s type with the @autoclosure attribute. Now you can call the
parameter is evaluated only in debug builds and its message parameter is evaluated function as if it took a String argument instead of a closure. The argument is
only if condition is false. automatically converted to a closure, because the customerProvider parameter’s type
is marked with the @autoclosure attribute.
An autoclosure lets you delay evaluation, because the code inside isn’t run until you
// customersInLine is ["Ewa", "Barry", "Daniella"]
call the closure. Delaying evaluation is useful for code that has side effects or is
func serve(customer customerProvider: @autoclosure () -> String) {

72
 print("Now serving \(customerProvider())!")
}
serve(customer: customersInLine.remove(at: 0))
// Prints "Now serving Ewa!"
NOTE

Overusing autoclosures can make your code hard to understand. The context and
function name should make it clear that evaluation is being deferred.

If you want an autoclosure that is allowed to escape, use both the @autoclosure and
@escaping attributes. The @escaping attribute is described above in Escaping
Closures.

// customersInLine is ["Barry", "Daniella"]

var customerProviders: [() -> String] = []
func collectCustomerProviders(_ customerProvider: @autoclosure
@escaping () -> String) {
customerProviders.append(customerProvider)
}
collectCustomerProviders(customersInLine.remove(at: 0))
collectCustomerProviders(customersInLine.remove(at: 0))

print("Collected \(customerProviders.count) closures.")

// Prints "Collected 2 closures."
for customerProvider in customerProviders {
print("Now serving \(customerProvider())!")
}
// Prints "Now serving Barry!"
// Prints "Now serving Daniella!"

In the code above, instead of calling the closure passed to it as its customerProvider
argument, the collectCustomerProviders(_:) function appends the closure to the
customerProviders array. The array is declared outside the scope of the function,
which means the closures in the array can be executed after the function returns. As a
result, the value of the customerProvider argument must be allowed to escape the
function’s scope.

73
 enum CompassPoint {
Section 8 case north

Enumerations case south

case east
case west
}

Enumerations The values defined in an enumeration (such as north, south, east, and west) are its
enumeration cases. You use the case keyword to introduce new enumeration cases.
An enumeration defines a common type for a group of related values and enables you
to work with those values in a type-safe way within your code. NOTE

Unlike C and Objective-C, Swift enumeration cases are not assigned a default integer
If you are familiar with C, you will know that C enumerations assign related names to value when they are created. In the CompassPoint example above, north, south, east
a set of integer values. Enumerations in Swift are much more flexible, and do not and west do not implicitly equal 0, 1, 2 and 3. Instead, the different enumeration cases
have to provide a value for each case of the enumeration. If a value (known as a “raw” are fully-fledged values in their own right, with an explicitly-defined type of
CompassPoint.
value) is provided for each enumeration case, the value can be a string, a character,
or a value of any integer or floating-point type. Multiple cases can appear on a single line, separated by commas:

Alternatively, enumeration cases can specify associated values of any type to be enum Planet {
stored along with each different case value, much as unions or variants do in other case mercury, venus, earth, mars, jupiter, saturn, uranus,
languages. You can define a common set of related cases as part of one neptune
enumeration, each of which has a different set of values of appropriate types }
associated with it.
Each enumeration definition defines a brand new type. Like other types in Swift, their
names (such as CompassPoint and Planet) should start with a capital letter. Give
Enumerations in Swift are first-class types in their own right. They adopt many
enumeration types singular rather than plural names, so that they read as self-
features traditionally supported only by classes, such as computed properties to
evident:
provide additional information about the enumeration’s current value, and instance
methods to provide functionality related to the values the enumeration represents. var directionToHead = CompassPoint.west
Enumerations can also define initializers to provide an initial case value; can be
extended to expand their functionality beyond their original implementation; and can The type of directionToHead is inferred when it is initialized with one of the possible
conform to protocols to provide standard functionality. values of CompassPoint. Once directionToHead is declared as a CompassPoint, you
can set it to a different CompassPoint value using a shorter dot syntax:
For more on these capabilities, see Properties, Methods, Initialization, Extensions,
directionToHead = .east
and Protocols.
The type of directionToHead is already known, and so you can drop the type when
Enumeration Syntax setting its value. This makes for highly readable code when working with explicitly-
typed enumeration values.
You introduce enumerations with the enum keyword and place their entire definition
within a pair of braces:
Matching Enumeration Values with a Switch Statement
enum SomeEnumeration { You can match individual enumeration values with a switch statement:
// enumeration definition goes here
directionToHead = .south
}
switch directionToHead {
Here’s an example for the four main points of a compass:
case .north:

74
 print("Lots of planets have a north") You can define Swift enumerations to store associated values of any given type, and
case .south: the value types can be different for each case of the enumeration if needed.
print("Watch out for penguins") Enumerations similar to these are known as discriminated unions, tagged unions, or
variants in other programming languages.
case .east:
print("Where the sun rises")
For example, suppose an inventory tracking system needs to track products by two
case .west: different types of barcode. Some products are labeled with 1D barcodes in UPC
print("Where the skies are blue") format, which uses the numbers 0 to 9. Each barcode has a “number system” digit,
} followed by five “manufacturer code” digits and five “product code” digits. These are
followed by a “check” digit to verify that the code has been scanned correctly:
// Prints "Watch out for penguins"

You can read this code as:

“Consider the value of directionToHead. In the case where it equals .north, print
"Lots of planets have a north". In the case where it equals .south, print "Watch
out for penguins".”

…and so on.

As described in Control Flow, a switch statement must be exhaustive when

considering an enumeration’s cases. If the case for .west is omitted, this code does
not compile, because it does not consider the complete list of CompassPoint cases.
Requiring exhaustiveness ensures that enumeration cases are not accidentally Other products are labeled with 2D barcodes in QR code format, which can use any
omitted. ISO 8859-1 character and can encode a string up to 2,953 characters long:

When it is not appropriate to provide a case for every enumeration case, you can
provide a default case to cover any cases that are not addressed explicitly:

let somePlanet = Planet.earth

switch somePlanet {
case .earth:
print("Mostly harmless")
default:
print("Not a safe place for humans")
}
// Prints "Mostly harmless"

Associated Values
The examples in the previous section show how the cases of an enumeration are a It would be convenient for an inventory tracking system to be able to store UPC
defined (and typed) value in their own right. You can set a constant or variable to barcodes as a tuple of four integers, and QR code barcodes as a string of any length.
Planet.earth, and check for this value later. However, it is sometimes useful to be
able to store associated values of other types alongside these case values. This In Swift, an enumeration to define product barcodes of either type might look like this:
enables you to store additional custom information along with the case value, and
permits this information to vary each time you use that case in your code.

75
 enum Barcode { If all of the associated values for an enumeration case are extracted as constants, or if
case upc(Int, Int, Int, Int) all are extracted as variables, you can place a single var or let annotation before the
case qrCode(String) case name, for brevity:
}
switch productBarcode {
This can be read as: case let .upc(numberSystem, manufacturer, product, check):
print("UPC : \(numberSystem), \(manufacturer), \(product), \
“Define an enumeration type called Barcode, which can take either a value of upc with (check).")
an associated value of type (Int, Int, Int, Int), or a value of qrCode with an case let .qrCode(productCode):
associated value of type String.”
print("QR code: \(productCode).")

This definition does not provide any actual Int or String values—it just defines the }
type of associated values that Barcode constants and variables can store when they // Prints "QR code: ABCDEFGHIJKLMNOP."
are equal to Barcode.upc or Barcode.qrCode.
Raw Values
New barcodes can then be created using either type: The barcode example in Associated Values shows how cases of an enumeration can
declare that they store associated values of different types. As an alternative to
var productBarcode = Barcode.upc(8, 85909, 51226, 3)
associated values, enumeration cases can come prepopulated with default values
This example creates a new variable called productBarcode and assigns it a value of (called raw values), which are all of the same type.
Barcode.upc with an associated tuple value of (8, 85909, 51226, 3).
Here’s an example that stores raw ASCII values alongside named enumeration cases:
The same product can be assigned a different type of barcode:
enum ASCIIControlCharacter: Character {
productBarcode = .qrCode("ABCDEFGHIJKLMNOP") case tab = "\t"
At this point, the original Barcode.upc and its integer values are replaced by the new case lineFeed = "\n"
Barcode.qrCode and its string value. Constants and variables of type Barcode can case carriageReturn = "\r"
store either a .upc or a .qrCode (together with their associated values), but they can
}
only store one of them at any given time.
Here, the raw values for an enumeration called ASCIIControlCharacter are defined to
The different barcode types can be checked using a switch statement, as before. This be of type Character, and are set to some of the more common ASCII control
time, however, the associated values can be extracted as part of the switch statement. characters. Character values are described in Strings and Characters.
You extract each associated value as a constant (with the let prefix) or a variable
(with the var prefix) for use within the switch case’s body: Raw values can be strings, characters, or any of the integer or floating-point number
types. Each raw value must be unique within its enumeration declaration.
switch productBarcode {
case .upc(let numberSystem, let manufacturer, let product, let NOTE
check): Raw values are not the same as associated values. Raw values are set to
print("UPC: \(numberSystem), \(manufacturer), \(product), \ prepopulated values when you first define the enumeration in your code, like the three
(check).") ASCII codes above. The raw value for a particular enumeration case is always the
case .qrCode(let productCode): same. Associated values are set when you create a new constant or variable based on
one of the enumeration’s cases, and can be different each time you do so.
print("QR code: \(productCode).")
} Implicitly Assigned Raw Values
// Prints "QR code: ABCDEFGHIJKLMNOP."

76
When you’re working with enumerations that store integer or string raw values, you This example identifies Uranus from its raw value of 7:
don’t have to explicitly assign a raw value for each case. When you don’t, Swift will
automatically assign the values for you. let possiblePlanet = Planet(rawValue: 7)
// possiblePlanet is of type Planet? and equals Planet.uranus
For instance, when integers are used for raw values, the implicit value for each case is
Not all possible Int values will find a matching planet, however. Because of this, the
one more than the previous case. If the first case doesn’t have a value set, its value is
raw value initializer always returns an optional enumeration case. In the example
0.
above, possiblePlanet is of type Planet?, or “optional Planet.”

The enumeration below is a refinement of the earlier Planet enumeration, with integer NOTE
raw values to represent each planet’s order from the sun:
The raw value initializer is a failable initializer, because not every raw value will return
an enumeration case. For more information, see Failable Initializers.
enum Planet: Int {
case mercury = 1, venus, earth, mars, jupiter, saturn, uranus, If you try to find a planet with a position of 11, the optional Planet value returned by
neptune the raw value initializer will be nil:
}
let positionToFind = 11
In the example above, Planet.mercury has an explicit raw value of 1, Planet.venus
has an implicit raw value of 2, and so on. if let somePlanet = Planet(rawValue: positionToFind) {
switch somePlanet {
When strings are used for raw values, the implicit value for each case is the text of case .earth:
that case’s name. print("Mostly harmless")
default:
The enumeration below is a refinement of the earlier CompassPoint enumeration, with
print("Not a safe place for humans")
string raw values to represent each direction’s name:
}
enum CompassPoint: String { } else {
case north, south, east, west print("There isn't a planet at position \(positionToFind)")
} }
In the example above, CompassPoint.south has an implicit raw value of "south", and // Prints "There isn't a planet at position 11"
so on. This example uses optional binding to try to access a planet with a raw value of 11.
The statement if let somePlanet = Planet(rawValue: 11) creates an optional
You access the raw value of an enumeration case with its rawValue property: Planet, and sets somePlanet to the value of that optional Planet if it can be retrieved.
In this case, it is not possible to retrieve a planet with a position of 11, and so the else
let earthsOrder = Planet.earth.rawValue
branch is executed instead.
// earthsOrder is 3

Recursive Enumerations
let sunsetDirection = CompassPoint.west.rawValue A recursive enumeration is an enumeration that has another instance of the
// sunsetDirection is "west" enumeration as the associated value for one or more of the enumeration cases. You
indicate that an enumeration case is recursive by writing indirect before it, which
Initializing from a Raw Value tells the compiler to insert the necessary layer of indirection.
If you define an enumeration with a raw-value type, the enumeration automatically
receives an initializer that takes a value of the raw value’s type (as a parameter called For example, here is an enumeration that stores simple arithmetic expressions:
rawValue) and returns either an enumeration case or nil. You can use this initializer
to try to create a new instance of the enumeration. enum ArithmeticExpression {

77
 case number(Int) }
indirect case addition(ArithmeticExpression,
ArithmeticExpression)
print(evaluate(product))
indirect case multiplication(ArithmeticExpression,
// Prints "18"
ArithmeticExpression)
} This function evaluates a plain number by simply returning the associated value. It
evaluates an addition or multiplication by evaluating the expression on the left hand
You can also write indirect before the beginning of the enumeration, to enable
side, evaluating the expression on the right hand side, and then adding them or
indirection for all of the enumeration’s cases that need it:
multiplying them.
indirect enum ArithmeticExpression {
case number(Int)
case addition(ArithmeticExpression, ArithmeticExpression)
case multiplication(ArithmeticExpression, ArithmeticExpression)
}

This enumeration can store three kinds of arithmetic expressions: a plain number, the
addition of two expressions, and the multiplication of two expressions. The addition
and multiplication cases have associated values that are also arithmetic
expressions—these associated values make it possible to nest expressions. For
example, the expression (5 + 4) * 2 has a number on the right hand side of the
multiplication and another expression on the left hand side of the multiplication.
Because the data is nested, the enumeration used to store the data also needs to
support nesting—this means the enumeration needs to be recursive. The code below
shows the ArithmeticExpression recursive enumeration being created for (5 + 4) *
2:

let five = ArithmeticExpression.number(5)

let four = ArithmeticExpression.number(4)
let sum = ArithmeticExpression.addition(five, four)
let product = ArithmeticExpression.multiplication(sum,
ArithmeticExpression.number(2))

A recursive function is a straightforward way to work with data that has a recursive
structure. For example, here’s a function that evaluates an arithmetic expression:

func evaluate(_ expression: ArithmeticExpression) -> Int {

switch expression {
case let .number(value):
return value
case let .addition(left, right):
return evaluate(left) + evaluate(right)
case let .multiplication(left, right):
return evaluate(left) * evaluate(right)
}

78
 Section 9 Deinitializers enable an instance of a class to free up any resources it has
assigned.

Classes and Structures Reference counting allows more than one reference to a class instance.
For more information, see Inheritance, Type Casting, Deinitialization, and Automatic
Reference Counting.

Classes and Structures NOTE

Structures are always copied when they are passed around in your code, and do not
Classes and structures are general-purpose, flexible constructs that become the use reference counting.
building blocks of your program’s code. You define properties and methods to add
Definition Syntax
functionality to your classes and structures by using exactly the same syntax as for
constants, variables, and functions. Classes and structures have a similar definition syntax. You introduce classes with the
class keyword and structures with the struct keyword. Both place their entire
Unlike other programming languages, Swift does not require you to create separate definition within a pair of braces:
interface and implementation files for custom classes and structures. In Swift, you
class SomeClass {
define a class or a structure in a single file, and the external interface to that class or
structure is automatically made available for other code to use. // class definition goes here
}
NOTE struct SomeStructure {
An instance of a class is traditionally known as an object. However, Swift classes and // structure definition goes here
structures are much closer in functionality than in other languages, and much of this
}
chapter describes functionality that can apply to instances of either a class or a
structure type. Because of this, the more general term instance is used. NOTE

Whenever you define a new class or structure, you effectively define a brand new Swift
Comparing Classes and Structures type. Give types UpperCamelCase names (such as SomeClass and SomeStructure
Classes and structures in Swift have many things in common. Both can: here) to match the capitalization of standard Swift types (such as String, Int, and
Bool). Conversely, always give properties and methods lowerCamelCase names (such
as frameRate and incrementCount) to differentiate them from type names.
Define properties to store values
Here’s an example of a structure definition and a class definition:
Define methods to provide functionality
Define subscripts to provide access to their values using subscript syntax struct Resolution {
var width = 0
Define initializers to set up their initial state
var height = 0
Be extended to expand their functionality beyond a default implementation
}
Conform to protocols to provide standard functionality of a certain kind
class VideoMode {
For more information, see Properties, Methods, Subscripts, Initialization, Extensions, var resolution = Resolution()
and Protocols.
var interlaced = false
var frameRate = 0.0
Classes have additional capabilities that structures do not:
var name: String?
Inheritance enables one class to inherit the characteristics of another. }

Type casting enables you to check and interpret the type of a class The example above defines a new structure called Resolution, to describe a pixel-
instance at runtime. based display resolution. This structure has two stored properties called width and

79
height. Stored properties are constants or variables that are bundled up and stored // Prints "The width of someVideoMode is 0"
as part of the class or structure. These two properties are inferred to be of type Int by You can also use dot syntax to assign a new value to a variable property:
setting them to an initial integer value of 0.
someVideoMode.resolution.width = 1280
The example above also defines a new class called VideoMode, to describe a specific print("The width of someVideoMode is now \
video mode for video display. This class has four variable stored properties. The first, (someVideoMode.resolution.width)")
resolution, is initialized with a new Resolution structure instance, which infers a // Prints "The width of someVideoMode is now 1280"
property type of Resolution. For the other three properties, new VideoMode instances
NOTE
will be initialized with an interlaced setting of false (meaning “noninterlaced video”),
a playback frame rate of 0.0, and an optional String value called name. The name Unlike Objective-C, Swift enables you to set sub-properties of a structure property
property is automatically given a default value of nil, or “no name value”, because it is directly. In the last example above, the width property of the resolution property of
someVideoMode is set directly, without your needing to set the entire resolution
of an optional type.
property to a new value.

Class and Structure Instances Memberwise Initializers for Structure Types

The Resolution structure definition and the VideoMode class definition only describe All structures have an automatically-generated memberwise initializer, which you can
what a Resolution or VideoMode will look like. They themselves do not describe a use to initialize the member properties of new structure instances. Initial values for the
specific resolution or video mode. To do that, you need to create an instance of the properties of the new instance can be passed to the memberwise initializer by name:
structure or class.
let vga = Resolution(width: 640, height: 480)
The syntax for creating instances is very similar for both structures and classes:
Unlike structures, class instances do not receive a default memberwise initializer.
Initializers are described in more detail in Initialization.
let someResolution = Resolution()
let someVideoMode = VideoMode()
Structures and Enumerations Are Value Types
Structures and classes both use initializer syntax for new instances. The simplest form
A value type is a type whose value is copied when it is assigned to a variable or
of initializer syntax uses the type name of the class or structure followed by empty
constant, or when it is passed to a function.
parentheses, such as Resolution() or VideoMode(). This creates a new instance of
the class or structure, with any properties initialized to their default values. Class and
You’ve actually been using value types extensively throughout the previous chapters.
structure initialization is described in more detail in Initialization.
In fact, all of the basic types in Swift—integers, floating-point numbers, Booleans,
strings, arrays and dictionaries—are value types, and are implemented as structures
Accessing Properties
behind the scenes.
You can access the properties of an instance using dot syntax. In dot syntax, you write
the property name immediately after the instance name, separated by a period (.), All structures and enumerations are value types in Swift. This means that any
without any spaces: structure and enumeration instances you create—and any value types they have as
properties—are always copied when they are passed around in your code.
print("The width of someResolution is \(someResolution.width)")
// Prints "The width of someResolution is 0" Consider this example, which uses the Resolution structure from the previous
In this example, someResolution.width refers to the width property of example:
someResolution, and returns its default initial value of 0.
let hd = Resolution(width: 1920, height: 1080)
You can drill down into sub-properties, such as the width property in the resolution var cinema = hd
property of a VideoMode: This example declares a constant called hd and sets it to a Resolution instance
initialized with the width and height of full HD video (1920 pixels wide by 1080 pixels
print("The width of someVideoMode is \
(someVideoMode.resolution.width)")
high).

80
It then declares a variable called cinema and sets it to the current value of hd. Unlike value types, reference types are not copied when they are assigned to a
Because Resolution is a structure, a copy of the existing instance is made, and this variable or constant, or when they are passed to a function. Rather than a copy, a
new copy is assigned to cinema. Even though hd and cinema now have the same reference to the same existing instance is used instead.
width and height, they are two completely different instances behind the scenes.
Here’s an example, using the VideoMode class defined above:
Next, the width property of cinema is amended to be the width of the slightly-wider 2K
standard used for digital cinema projection (2048 pixels wide and 1080 pixels high): let tenEighty = VideoMode()
tenEighty.resolution = hd
cinema.width = 2048
tenEighty.interlaced = true
Checking the width property of cinema shows that it has indeed changed to be 2048: tenEighty.name = "1080i"
tenEighty.frameRate = 25.0
print("cinema is now \(cinema.width) pixels wide")
// Prints "cinema is now 2048 pixels wide" This example declares a new constant called tenEighty and sets it to refer to a new
instance of the VideoMode class. The video mode is assigned a copy of the HD
However, the width property of the original hd instance still has the old value of 1920: resolution of 1920 by 1080 from before. It is set to be interlaced, and is given a name
of "1080i". Finally, it is set to a frame rate of 25.0 frames per second.
print("hd is still \(hd.width) pixels wide")
// Prints "hd is still 1920 pixels wide" Next, tenEighty is assigned to a new constant, called alsoTenEighty, and the frame
When cinema was given the current value of hd, the values stored in hd were copied rate of alsoTenEighty is modified:
into the new cinema instance. The end result is two completely separate instances,
which just happened to contain the same numeric values. Because they are separate let alsoTenEighty = tenEighty
instances, setting the width of cinema to 2048 doesn’t affect the width stored in hd. alsoTenEighty.frameRate = 30.0

Because classes are reference types, tenEighty and alsoTenEighty actually both
The same behavior applies to enumerations: refer to the same VideoMode instance. Effectively, they are just two different names for
the same single instance.
enum CompassPoint {
case north, south, east, west Checking the frameRate property of tenEighty shows that it correctly reports the new
} frame rate of 30.0 from the underlying VideoMode instance:
var currentDirection = CompassPoint.west
print("The frameRate property of tenEighty is now \
let rememberedDirection = currentDirection
(tenEighty.frameRate)")
currentDirection = .east
// Prints "The frameRate property of tenEighty is now 30.0"
if rememberedDirection == .west {
Note that tenEighty and alsoTenEighty are declared as constants, rather than
print("The remembered direction is still .west") variables. However, you can still change tenEighty.frameRate and
} alsoTenEighty.frameRate because the values of the tenEighty and alsoTenEighty
// Prints "The remembered direction is still .west" constants themselves do not actually change. tenEighty and alsoTenEighty
themselves do not “store” the VideoMode instance—instead, they both refer to a
When rememberedDirection is assigned the value of currentDirection, it is actually
VideoMode instance behind the scenes. It is the frameRate property of the underlying
set to a copy of that value. Changing the value of currentDirection thereafter does
VideoMode that is changed, not the values of the constant references to that
not affect the copy of the original value that was stored in rememberedDirection.
VideoMode.

Classes Are Reference Types Identity Operators

Because classes are reference types, it is possible for multiple constants and
variables to refer to the same single instance of a class behind the scenes. (The same

81
is not true for structures and enumerations, because they are always copied when However, structure instances are always passed by value, and class instances are
they are assigned to a constant or variable, or passed to a function.) always passed by reference. This means that they are suited to different kinds of
tasks. As you consider the data constructs and functionality that you need for a
It can sometimes be useful to find out if two constants or variables refer to exactly the project, decide whether each data construct should be defined as a class or as a
same instance of a class. To enable this, Swift provides two identity operators: structure.

Identical to (===) As a general guideline, consider creating a structure when one or more of these
conditions apply:
Not identical to (!==)
Use these operators to check whether two constants or variables refer to the same The structure’s primary purpose is to encapsulate a few relatively simple
single instance: data values.

if tenEighty === alsoTenEighty { It is reasonable to expect that the encapsulated values will be copied rather
than referenced when you assign or pass around an instance of that
print("tenEighty and alsoTenEighty refer to the same VideoMode
instance.") structure.
} Any properties stored by the structure are themselves value types, which
// Prints "tenEighty and alsoTenEighty refer to the same VideoMode would also be expected to be copied rather than referenced.
instance." The structure does not need to inherit properties or behavior from another
Note that “identical to” (represented by three equals signs, or ===) does not mean the existing type.
same thing as “equal to” (represented by two equals signs, or ==): Examples of good candidates for structures include:

“Identical to” means that two constants or variables of class type refer to The size of a geometric shape, perhaps encapsulating a width property
exactly the same class instance. and a height property, both of type Double.
“Equal to” means that two instances are considered “equal” or “equivalent” A way to refer to ranges within a series, perhaps encapsulating a start
in value, for some appropriate meaning of “equal”, as defined by the type’s property and a length property, both of type Int.
designer.
A point in a 3D coordinate system, perhaps encapsulating x, y and z
When you define your own custom classes and structures, it is your responsibility to properties, each of type Double.
decide what qualifies as two instances being “equal”. The process of defining your
own implementations of the “equal to” and “not equal to” operators is described in In all other cases, define a class, and create instances of that class to be managed
Equivalence Operators. and passed by reference. In practice, this means that most custom data constructs
should be classes, not structures.
Pointers
If you have experience with C, C++, or Objective-C, you may know that these Assignment and Copy Behavior for Strings, Arrays, and
languages use pointers to refer to addresses in memory. A Swift constant or variable Dictionaries
that refers to an instance of some reference type is similar to a pointer in C, but is not In Swift, many basic data types such as String, Array, and Dictionary are
a direct pointer to an address in memory, and does not require you to write an asterisk implemented as structures. This means that data such as strings, arrays, and
(*) to indicate that you are creating a reference. Instead, these references are defined dictionaries are copied when they are assigned to a new constant or variable, or when
like any other constant or variable in Swift. they are passed to a function or method.

Choosing Between Classes and Structures This behavior is different from Foundation: NSString, NSArray, and NSDictionary are
You can use both classes and structures to define custom data types to use as the implemented as classes, not structures. Strings, arrays, and dictionaries in Foundation
building blocks of your program’s code. are always assigned and passed around as a reference to an existing instance, rather
than as a copy.

82
NOTE

The description above refers to the “copying” of strings, arrays, and dictionaries. The
behavior you see in your code will always be as if a copy took place. However, Swift
only performs an actual copy behind the scenes when it is absolutely necessary to do
so. Swift manages all value copying to ensure optimal performance, and you should not
avoid assignment to try to preempt this optimization.

83
 rangeOfThreeItems.firstValue = 6
Section 10 // the range now represents integer values 6, 7, and 8

Properties Instances of FixedLengthRange have a variable stored property called firstValue

and a constant stored property called length. In the example above, length is
initialized when the new range is created and cannot be changed thereafter, because
it is a constant property.

Properties Stored Properties of Constant Structure Instances

If you create an instance of a structure and assign that instance to a constant, you
Properties associate values with a particular class, structure, or enumeration. Stored cannot modify the instance’s properties, even if they were declared as variable
properties store constant and variable values as part of an instance, whereas properties:
computed properties calculate (rather than store) a value. Computed properties are
provided by classes, structures, and enumerations. Stored properties are provided let rangeOfFourItems = FixedLengthRange(firstValue: 0, length: 4)
only by classes and structures.
// this range represents integer values 0, 1, 2, and 3
rangeOfFourItems.firstValue = 6
Stored and computed properties are usually associated with instances of a particular
type. However, properties can also be associated with the type itself. Such properties // this will report an error, even though firstValue is a variable
property
are known as type properties.
Because rangeOfFourItems is declared as a constant (with the let keyword), it is not
In addition, you can define property observers to monitor changes in a property’s possible to change its firstValue property, even though firstValue is a variable
value, which you can respond to with custom actions. Property observers can be property.
added to stored properties you define yourself, and also to properties that a subclass
inherits from its superclass. This behavior is due to structures being value types. When an instance of a value
type is marked as a constant, so are all of its properties.
Stored Properties
The same is not true for classes, which are reference types. If you assign an instance
In its simplest form, a stored property is a constant or variable that is stored as part of
of a reference type to a constant, you can still change that instance’s variable
an instance of a particular class or structure. Stored properties can be either variable
properties.
stored properties (introduced by the var keyword) or constant stored properties
(introduced by the let keyword).
Lazy Stored Properties
You can provide a default value for a stored property as part of its definition, as A lazy stored property is a property whose initial value is not calculated until the first
described in Default Property Values. You can also set and modify the initial value for time it is used. You indicate a lazy stored property by writing the lazy modifier before
a stored property during initialization. This is true even for constant stored properties, its declaration.
as described in Assigning Constant Properties During Initialization.
NOTE

The example below defines a structure called FixedLengthRange, which describes a You must always declare a lazy property as a variable (with the var keyword), because
range of integers whose range length cannot be changed once it is created: its initial value might not be retrieved until after instance initialization completes.
Constant properties must always have a value before initialization completes, and
struct FixedLengthRange { therefore cannot be declared as lazy.

var firstValue: Int Lazy properties are useful when the initial value for a property is dependent on
let length: Int outside factors whose values are not known until after an instance’s initialization is
}
complete. Lazy properties are also useful when the initial value for a property requires
complex or computationally expensive setup that should not be performed unless or
var rangeOfThreeItems = FixedLengthRange(firstValue: 0, length: 3)
until it is needed.
// the range represents integer values 0, 1, and 2

84
The example below uses a lazy stored property to avoid unnecessary initialization of a Because it is marked with the lazy modifier, the DataImporter instance for the
complex class. This example defines two classes called DataImporter and importer property is only created when the importer property is first accessed, such
DataManager, neither of which is shown in full: as when its fileName property is queried:

class DataImporter { print(manager.importer.fileName)

/* // the DataImporter instance for the importer property has now been
created
DataImporter is a class to import data from an external file.
// Prints "data.txt"
The class is assumed to take a non-trivial amount of time to
initialize. NOTE
*/ If a property marked with the lazy modifier is accessed by multiple threads
var fileName = "data.txt" simultaneously and the property has not yet been initialized, there is no guarantee that
the property will be initialized only once.
// the DataImporter class would provide data importing
functionality here
Stored Properties and Instance Variables
}
If you have experience with Objective-C, you may know that it provides two ways to
store values and references as part of a class instance. In addition to properties, you
class DataManager { can use instance variables as a backing store for the values stored in a property.
lazy var importer = DataImporter()
var data = [String]() Swift unifies these concepts into a single property declaration. A Swift property does
not have a corresponding instance variable, and the backing store for a property is not
// the DataManager class would provide data management
functionality here accessed directly. This approach avoids confusion about how the value is accessed in
different contexts and simplifies the property’s declaration into a single, definitive
}
statement. All information about the property—including its name, type, and memory
management characteristics—is defined in a single location as part of the type’s
let manager = DataManager() definition.
manager.data.append("Some data")
manager.data.append("Some more data") Computed Properties
// the DataImporter instance for the importer property has not yet In addition to stored properties, classes, structures, and enumerations can define
been created computed properties, which do not actually store a value. Instead, they provide a
The DataManager class has a stored property called data, which is initialized with a getter and an optional setter to retrieve and set other properties and values indirectly.
new, empty array of String values. Although the rest of its functionality is not shown,
the purpose of this DataManager class is to manage and provide access to this array of struct Point {
String data. var x = 0.0, y = 0.0
}
Part of the functionality of the DataManager class is the ability to import data from a struct Size {
file. This functionality is provided by the DataImporter class, which is assumed to take
var width = 0.0, height = 0.0
a non-trivial amount of time to initialize. This might be because a DataImporter
instance needs to open a file and read its contents into memory when the }
DataImporter instance is initialized. struct Rect {
var origin = Point()
It is possible for a DataManager instance to manage its data without ever importing var size = Size()
data from a file, so there is no need to create a new DataImporter instance when the
var center: Point {
DataManager itself is created. Instead, it makes more sense to create the
DataImporter instance if and when it is first used. get {

85
 let centerX = origin.x + (size.width / 2) below. Setting the center property calls the setter for center, which modifies the x
let centerY = origin.y + (size.height / 2) and y values of the stored origin property, and moves the square to its new position.
return Point(x: centerX, y: centerY)
}
set(newCenter) {
origin.x = newCenter.x - (size.width / 2)
origin.y = newCenter.y - (size.height / 2)
}
}
}
var square = Rect(origin: Point(x: 0.0, y: 0.0),
size: Size(width: 10.0, height: 10.0))
let initialSquareCenter = square.center
square.center = Point(x: 15.0, y: 15.0)
print("square.origin is now at (\(square.origin.x), \
(square.origin.y))")
// Prints "square.origin is now at (10.0, 10.0)"

This example defines three structures for working with geometric shapes:

Point encapsulates the x- and y-coordinate of a point.

Size encapsulates a width and a height. Shorthand Setter Declaration

Rect defines a rectangle by an origin point and a size. If a computed property’s setter does not define a name for the new value to be set, a
The Rect structure also provides a computed property called center. The current default name of newValue is used. Here’s an alternative version of the Rect structure,
center position of a Rect can always be determined from its origin and size, and so which takes advantage of this shorthand notation:
you don’t need to store the center point as an explicit Point value. Instead, Rect
struct AlternativeRect {
defines a custom getter and setter for a computed variable called center, to enable
you to work with the rectangle’s center as if it were a real stored property. var origin = Point()
var size = Size()
The preceding example creates a new Rect variable called square. The square var center: Point {
variable is initialized with an origin point of (0, 0), and a width and height of 10. This get {
square is represented by the blue square in the diagram below.
let centerX = origin.x + (size.width / 2)

The square variable’s center property is then accessed through dot syntax let centerY = origin.y + (size.height / 2)
(square.center), which causes the getter for center to be called, to retrieve the return Point(x: centerX, y: centerY)
current property value. Rather than returning an existing value, the getter actually }
calculates and returns a new Point to represent the center of the square. As can be set {
seen above, the getter correctly returns a center point of (5, 5).
origin.x = newValue.x - (size.width / 2)

The center property is then set to a new value of (15, 15), which moves the square origin.y = newValue.y - (size.height / 2)
up and to the right, to the new position shown by the orange square in the diagram }
}

86
 } you can observe and respond to changes to their value in the computed property’s
setter. Property overriding is described in Overriding.
Read-Only Computed Properties
A computed property with a getter but no setter is known as a read-only computed You have the option to define either or both of these observers on a property:
property. A read-only computed property always returns a value, and can be accessed
through dot syntax, but cannot be set to a different value. willSet is called just before the value is stored.

NOTE didSet is called immediately after the new value is stored.

You must declare computed properties—including read-only computed properties—as If you implement a willSet observer, it’s passed the new property value as a constant
variable properties with the var keyword, because their value is not fixed. The let parameter. You can specify a name for this parameter as part of your willSet
keyword is only used for constant properties, to indicate that their values cannot be implementation. If you don’t write the parameter name and parentheses within your
changed once they are set as part of instance initialization. implementation, the parameter is made available with a default parameter name of
You can simplify the declaration of a read-only computed property by removing the newValue.
get keyword and its braces:
Similarly, if you implement a didSet observer, it’s passed a constant parameter
struct Cuboid { containing the old property value. You can name the parameter or use the default
var width = 0.0, height = 0.0, depth = 0.0 parameter name of oldValue. If you assign a value to a property within its own didSet
observer, the new value that you assign replaces the one that was just set.
var volume: Double {
return width * height * depth NOTE
} The willSet and didSet observers of superclass properties are called when a
} property is set in a subclass initializer, after the superclass initializer has been called.
They are not called while a class is setting its own properties, before the superclass
let fourByFiveByTwo = Cuboid(width: 4.0, height: 5.0, depth: 2.0)
initializer has been called.
print("the volume of fourByFiveByTwo is \(fourByFiveByTwo.volume)")
For more information about initializer delegation, see Initializer Delegation for Value
// Prints "the volume of fourByFiveByTwo is 40.0" Types and Initializer Delegation for Class Types.
This example defines a new structure called Cuboid, which represents a 3D
Here’s an example of willSet and didSet in action. The example below defines a
rectangular box with width, height, and depth properties. This structure also has a
new class called StepCounter, which tracks the total number of steps that a person
read-only computed property called volume, which calculates and returns the current
takes while walking. This class might be used with input data from a pedometer or
volume of the cuboid. It doesn’t make sense for volume to be settable, because it
other step counter to keep track of a person’s exercise during their daily routine.
would be ambiguous as to which values of width, height, and depth should be used
for a particular volume value. Nonetheless, it is useful for a Cuboid to provide a read- class StepCounter {
only computed property to enable external users to discover its current calculated
var totalSteps: Int = 0 {
volume.
willSet(newTotalSteps) {

Property Observers print("About to set totalSteps to \(newTotalSteps)")

Property observers observe and respond to changes in a property’s value. Property }

observers are called every time a property’s value is set, even if the new value is the didSet {
same as the property’s current value. if totalSteps > oldValue {
print("Added \(totalSteps - oldValue) steps")
You can add property observers to any stored properties you define, except for lazy
}
stored properties. You can also add property observers to any inherited property
(whether stored or computed) by overriding the property within a subclass. You don’t }
need to define property observers for nonoverridden computed properties, because }
}

87
 let stepCounter = StepCounter() However, you can also define computed variables and define observers for stored
stepCounter.totalSteps = 200 variables, in either a global or local scope. Computed variables calculate their value,
// About to set totalSteps to 200 rather than storing it, and they are written in the same way as computed properties.
// Added 200 steps
NOTE
stepCounter.totalSteps = 360
Global constants and variables are always computed lazily, in a similar manner to Lazy
// About to set totalSteps to 360 Stored Properties. Unlike lazy stored properties, global constants and variables do not
// Added 160 steps need to be marked with the lazy modifier.
stepCounter.totalSteps = 896 Local constants and variables are never computed lazily.
// About to set totalSteps to 896
// Added 536 steps Type Properties
The StepCounter class declares a totalSteps property of type Int. This is a stored Instance properties are properties that belong to an instance of a particular type.
property with willSet and didSet observers. Every time you create a new instance of that type, it has its own set of property
values, separate from any other instance.
The willSet and didSet observers for totalSteps are called whenever the property
is assigned a new value. This is true even if the new value is the same as the current You can also define properties that belong to the type itself, not to any one instance of
value. that type. There will only ever be one copy of these properties, no matter how many
instances of that type you create. These kinds of properties are called type properties.
This example’s willSet observer uses a custom parameter name of newTotalSteps
for the upcoming new value. In this example, it simply prints out the value that is about Type properties are useful for defining values that are universal to all instances of a
to be set. particular type, such as a constant property that all instances can use (like a static
constant in C), or a variable property that stores a value that is global to all instances
The didSet observer is called after the value of totalSteps is updated. It compares of that type (like a static variable in C).
the new value of totalSteps against the old value. If the total number of steps has
increased, a message is printed to indicate how many new steps have been taken. Stored type properties can be variables or constants. Computed type properties are
The didSet observer does not provide a custom parameter name for the old value, always declared as variable properties, in the same way as computed instance
and the default name of oldValue is used instead. properties.

NOTE NOTE

If you pass a property that has observers to a function as an in-out parameter, the Unlike stored instance properties, you must always give stored type properties a default
willSet and didSet observers are always called. This is because of the copy-in copy- value. This is because the type itself does not have an initializer that can assign a value
out memory model for in-out parameters: The value is always written back to the to a stored type property at initialization time.
property at the end of the function. For a detailed discussion of the behavior of in-out Stored type properties are lazily initialized on their first access. They are guaranteed to
parameters, see In-Out Parameters. be initialized only once, even when accessed by multiple threads simultaneously, and
they do not need to be marked with the lazy modifier.
Global and Local Variables
The capabilities described above for computing and observing properties are also
Type Property Syntax
available to global variables and local variables. Global variables are variables that In C and Objective-C, you define static constants and variables associated with a type
are defined outside of any function, method, closure, or type context. Local variables as global static variables. In Swift, however, type properties are written as part of the
are variables that are defined within a function, method, or closure context. type’s definition, within the type’s outer curly braces, and each type property is
explicitly scoped to the type it supports.
The global and local variables you have encountered in previous chapters have all
been stored variables. Stored variables, like stored properties, provide storage for a You define type properties with the static keyword. For computed type properties for
value of a certain type and allow that value to be set and retrieved. class types, you can use the class keyword instead to allow subclasses to override

88
the superclass’s implementation. The example below shows the syntax for stored and // Prints "6"
computed type properties: print(SomeClass.computedTypeProperty)
// Prints "27"
struct SomeStructure {
The examples that follow use two stored type properties as part of a structure that
static var storedTypeProperty = "Some value."
models an audio level meter for a number of audio channels. Each channel has an
static var computedTypeProperty: Int { integer audio level between 0 and 10 inclusive.
return 1
} The figure below illustrates how two of these audio channels can be combined to
} model a stereo audio level meter. When a channel’s audio level is 0, none of the lights
for that channel are lit. When the audio level is 10, all of the lights for that channel are
enum SomeEnumeration {
lit. In this figure, the left channel has a current level of 9, and the right channel has a
static var storedTypeProperty = "Some value." current level of 7:
static var computedTypeProperty: Int {
return 6
}
}
class SomeClass {
static var storedTypeProperty = "Some value."
static var computedTypeProperty: Int {
return 27
}
class var overrideableComputedTypeProperty: Int {
return 107
}
}
NOTE

The computed type property examples above are for read-only computed type
properties, but you can also define read-write computed type properties with the same
syntax as for computed instance properties.

Querying and Setting Type Properties

Type properties are queried and set with dot syntax, just like instance properties.
However, type properties are queried and set on the type, not on an instance of that
type. For example:

print(SomeStructure.storedTypeProperty)
// Prints "Some value."
SomeStructure.storedTypeProperty = "Another value."
print(SomeStructure.storedTypeProperty) The audio channels described above are represented by instances of the
AudioChannel structure:
// Prints "Another value."
print(SomeEnumeration.computedTypeProperty) struct AudioChannel {

89
 static let thresholdLevel = 10 In the first of these two checks, the didSet observer sets currentLevel to a different
static var maxInputLevelForAllChannels = 0 value. This does not, however, cause the observer to be called again.

var currentLevel: Int = 0 { You can use the AudioChannel structure to create two new audio channels called
didSet { leftChannel and rightChannel, to represent the audio levels of a stereo sound

if currentLevel > AudioChannel.thresholdLevel {

system:
// cap the new audio level to the threshold level var leftChannel = AudioChannel()
currentLevel = AudioChannel.thresholdLevel var rightChannel = AudioChannel()
}
If you set the currentLevel of the left channel to 7, you can see that the
if currentLevel > maxInputLevelForAllChannels type property is updated to equal 7:
AudioChannel.maxInputLevelForAllChannels {
// store this as the new overall maximum input level leftChannel.currentLevel = 7
AudioChannel.maxInputLevelForAllChannels = print(leftChannel.currentLevel)
currentLevel
// Prints "7"
}
print(AudioChannel.maxInputLevelForAllChannels)
}
// Prints "7"
}
If you try to set the currentLevel of the right channel to 11, you can see that the right
}
channel’s currentLevel property is capped to the maximum value of 10, and the
The AudioChannel structure defines two stored type properties to support its maxInputLevelForAllChannels type property is updated to equal 10:
functionality. The first, thresholdLevel, defines the maximum threshold value an
audio level can take. This is a constant value of 10 for all AudioChannel instances. If rightChannel.currentLevel = 11
an audio signal comes in with a higher value than 10, it will be capped to this threshold print(rightChannel.currentLevel)
value (as described below).
// Prints "10"
print(AudioChannel.maxInputLevelForAllChannels)
The second type property is a variable stored property called
maxInputLevelForAllChannels. This keeps track of the maximum input value that has // Prints "10"
been received by any AudioChannel instance. It starts with an initial value of 0.

The AudioChannel structure also defines a stored instance property called

currentLevel, which represents the channel’s current audio level on a scale of 0 to
10.

The currentLevel property has a didSet property observer to check the value of
currentLevel whenever it is set. This observer performs two checks:

If the new value of currentLevel is greater than the allowed

thresholdLevel, the property observer caps currentLevel to
thresholdLevel.

If the new value of currentLevel (after any capping) is higher than any
value previously received by any AudioChannel instance, the property
observer stores the new currentLevel value in the
maxInputLevelForAllChannels type property.
NOTE

90
 func reset() {
Section 11 count = 0

Methods }
}

The Counter class defines three instance methods:

Methods increment() increments the counter by 1.

increment(by: Int) increments the counter by a specified integer

Methods are functions that are associated with a particular type. Classes, structures, amount.
and enumerations can all define instance methods, which encapsulate specific tasks
reset() resets the counter to zero.
and functionality for working with an instance of a given type. Classes, structures, and
enumerations can also define type methods, which are associated with the type itself. The Counter class also declares a variable property, count, to keep track of the
Type methods are similar to class methods in Objective-C. current counter value.

The fact that structures and enumerations can define methods in Swift is a major You call instance methods with the same dot syntax as properties:
difference from C and Objective-C. In Objective-C, classes are the only types that can
define methods. In Swift, you can choose whether to define a class, structure, or let counter = Counter()
enumeration, and still have the flexibility to define methods on the type you create. // the initial counter value is 0
counter.increment()
Instance Methods // the counter's value is now 1
Instance methods are functions that belong to instances of a particular class, counter.increment(by: 5)
structure, or enumeration. They support the functionality of those instances, either by // the counter's value is now 6
providing ways to access and modify instance properties, or by providing functionality
counter.reset()
related to the instance’s purpose. Instance methods have exactly the same syntax as
functions, as described in Functions. // the counter's value is now 0

Function parameters can have both a name (for use within the function’s body) and an
You write an instance method within the opening and closing braces of the type it argument label (for use when calling the function), as described in Function Argument
belongs to. An instance method has implicit access to all other instance methods and Labels and Parameter Names. The same is true for method parameters, because
properties of that type. An instance method can be called only on a specific instance methods are just functions that are associated with a type.
of the type it belongs to. It cannot be called in isolation without an existing instance.
The self Property
Here’s an example that defines a simple Counter class, which can be used to count Every instance of a type has an implicit property called self, which is exactly
the number of times an action occurs: equivalent to the instance itself. You use the self property to refer to the current
instance within its own instance methods.
class Counter {
var count = 0 The increment() method in the example above could have been written like this:
func increment() {
count += 1 func increment() {

} self.count += 1

func increment(by amount: Int) { }

count += amount In practice, you don’t need to write self in your code very often. If you don’t explicitly
} write self, Swift assumes that you are referring to a property or method of the current

91
instance whenever you use a known property or method name within a method. This mutating func moveBy(x deltaX: Double, y deltaY: Double) {
assumption is demonstrated by the use of count (rather than self.count) inside the x += deltaX
three instance methods for Counter. y += deltaY
}
The main exception to this rule occurs when a parameter name for an instance
}
method has the same name as a property of that instance. In this situation, the
parameter name takes precedence, and it becomes necessary to refer to the property var somePoint = Point(x: 1.0, y: 1.0)
in a more qualified way. You use the self property to distinguish between the somePoint.moveBy(x: 2.0, y: 3.0)
parameter name and the property name. print("The point is now at (\(somePoint.x), \(somePoint.y))")
// Prints "The point is now at (3.0, 4.0)"
Here, self disambiguates between a method parameter called x and an instance
property that is also called x: The Point structure above defines a mutating moveBy(x:y:) method, which moves a
Point instance by a certain amount. Instead of returning a new point, this method
struct Point { actually modifies the point on which it is called. The mutating keyword is added to its
var x = 0.0, y = 0.0 definition to enable it to modify its properties.
func isToTheRightOf(x: Double) -> Bool {
Note that you cannot call a mutating method on a constant of structure type, because
return self.x > x its properties cannot be changed, even if they are variable properties, as described in
} Stored Properties of Constant Structure Instances:
}
let fixedPoint = Point(x: 3.0, y: 3.0)
let somePoint = Point(x: 4.0, y: 5.0)
fixedPoint.moveBy(x: 2.0, y: 3.0)
if somePoint.isToTheRightOf(x: 1.0) {
// this will report an error
print("This point is to the right of the line where x == 1.0")
} Assigning to self Within a Mutating Method
// Prints "This point is to the right of the line where x == 1.0" Mutating methods can assign an entirely new instance to the implicit self property.
Without the self prefix, Swift would assume that both uses of x referred to the method The Point example shown above could have been written in the following way
parameter called x. instead:

struct Point {
Modifying Value Types from Within Instance Methods
var x = 0.0, y = 0.0
Structures and enumerations are value types. By default, the properties of a value
type cannot be modified from within its instance methods. mutating func moveBy(x deltaX: Double, y deltaY: Double) {
self = Point(x: x + deltaX, y: y + deltaY)
However, if you need to modify the properties of your structure or enumeration within }
a particular method, you can opt in to mutating behavior for that method. The method
}
can then mutate (that is, change) its properties from within the method, and any
changes that it makes are written back to the original structure when the method ends. This version of the mutating moveBy(x:y:) method creates a brand new structure
The method can also assign a completely new instance to its implicit self property, whose x and y values are set to the target location. The end result of calling this
and this new instance will replace the existing one when the method ends. alternative version of the method will be exactly the same as for calling the earlier
version.
You can opt in to this behavior by placing the mutating keyword before the func
keyword for that method: Mutating methods for enumerations can set the implicit self parameter to be a
different case from the same enumeration:
struct Point {
enum TriStateSwitch {
var x = 0.0, y = 0.0

92
 case off, low, high }
mutating func next() { }
switch self { SomeClass.someTypeMethod()
case .off: Within the body of a type method, the implicit self property refers to the type itself,
self = .low rather than an instance of that type. This means that you can use self to
case .low: disambiguate between type properties and type method parameters, just as you do for
instance properties and instance method parameters.
self = .high
case .high:
More generally, any unqualified method and property names that you use within the
self = .off body of a type method will refer to other type-level methods and properties. A type
} method can call another type method with the other method’s name, without needing
} to prefix it with the type name. Similarly, type methods on structures and enumerations
can access type properties by using the type property’s name without a type name
}
prefix.
var ovenLight = TriStateSwitch.low
ovenLight.next() The example below defines a structure called LevelTracker, which tracks a player’s
// ovenLight is now equal to .high progress through the different levels or stages of a game. It is a single-player game,
ovenLight.next() but can store information for multiple players on a single device.
// ovenLight is now equal to .off
All of the game’s levels (apart from level one) are locked when the game is first
This example defines an enumeration for a three-state switch. The switch cycles played. Every time a player finishes a level, that level is unlocked for all players on the
between three different power states (off, low and high) every time its next() method device. The LevelTracker structure uses type properties and methods to keep track of
is called. which levels of the game have been unlocked. It also tracks the current level for an
individual player.
Type Methods
struct LevelTracker {
Instance methods, as described above, are methods that are called on an instance of
a particular type. You can also define methods that are called on the type itself. These static var highestUnlockedLevel = 1
kinds of methods are called type methods. You indicate type methods by writing the var currentLevel = 1
static keyword before the method’s func keyword. Classes may also use the class
keyword to allow subclasses to override the superclass’s implementation of that
static func unlock(_ level: Int) {
method.
if level > highestUnlockedLevel { highestUnlockedLevel =
level }
NOTE
}
In Objective-C, you can define type-level methods only for Objective-C classes. In
Swift, you can define type-level methods for all classes, structures, and enumerations.
Each type method is explicitly scoped to the type it supports. static func isUnlocked(_ level: Int) -> Bool {

Type methods are called with dot syntax, like instance methods. However, you call return level <= highestUnlockedLevel
type methods on the type, not on an instance of that type. Here’s how you call a type }
method on a class called SomeClass:
@discardableResult
class SomeClass {
mutating func advance(to level: Int) -> Bool {
class func someTypeMethod() {
if LevelTracker.isUnlocked(level) {
// type method implementation goes here
currentLevel = level

93
 return true }
} else { }
return false The Player class creates a new instance of LevelTracker to track that player’s
} progress. It also provides a method called complete(level:), which is called
} whenever a player completes a particular level. This method unlocks the next level for
all players and updates the player’s progress to move them to the next level. (The
}
Boolean return value of advance(to:) is ignored, because the level is known to have
The LevelTracker structure keeps track of the highest level that any player has been unlocked by the call to LevelTracker.unlock(_:) on the previous line.)
unlocked. This value is stored in a type property called highestUnlockedLevel.
You can create an instance of the Player class for a new player, and see what
LevelTracker also defines two type functions to work with the highestUnlockedLevel happens when the player completes level one:
property. The first is a type function called unlock(_:), which updates the value of
highestUnlockedLevel whenever a new level is unlocked. The second is a var player = Player(name: "Argyrios")
convenience type function called isUnlocked(_:), which returns true if a particular player.complete(level: 1)
level number is already unlocked. (Note that these type methods can access the
print("highest unlocked level is now \
highestUnlockedLevel type property without your needing to write it as (LevelTracker.highestUnlockedLevel)")
LevelTracker.highestUnlockedLevel.)
// Prints "highest unlocked level is now 2"

In addition to its type property and type methods, LevelTracker tracks an individual If you create a second player, whom you try to move to a level that is not yet unlocked
player’s progress through the game. It uses an instance property called currentLevel by any player in the game, the attempt to set the player’s current level fails:
to track the level that a player is currently playing.
player = Player(name: "Beto")

To help manage the currentLevel property, LevelTracker defines an instance if player.tracker.advance(to: 6) {

method called advance(to:). Before updating currentLevel, this method checks print("player is now on level 6")
whether the requested new level is already unlocked. The advance(to:) method } else {
returns a Boolean value to indicate whether or not it was actually able to set
print("level 6 has not yet been unlocked")
currentLevel. Because it’s not necessarily a mistake for code that calls the
advance(to:) method to ignore the return value, this function is marked with the }
@discardableResult attribute. For more information about this attribute, see // Prints "level 6 has not yet been unlocked"
Attributes.

The LevelTracker structure is used with the Player class, shown below, to track and
update the progress of an individual player:

class Player {
var tracker = LevelTracker()
let playerName: String
func complete(level: Int) {
LevelTracker.unlock(level + 1)
tracker.advance(to: level + 1)
}
init(name: String) {
playerName = name

94
 Section 12 As with read-only computed properties, you can drop the get keyword for read-only
subscripts:

Subscripts subscript(index: Int) -> Int {

// return an appropriate subscript value here
}

Here’s an example of a read-only subscript implementation, which defines a

Subscripts TimesTable structure to represent an n-times-table of integers:

Classes, structures, and enumerations can define subscripts, which are shortcuts for struct TimesTable {
accessing the member elements of a collection, list, or sequence. You use subscripts
let multiplier: Int
to set and retrieve values by index without needing separate methods for setting and
retrieval. For example, you access elements in an Array instance as subscript(index: Int) -> Int {
someArray[index] and elements in a Dictionary instance as someDictionary[key]. return multiplier * index
}
You can define multiple subscripts for a single type, and the appropriate subscript }
overload to use is selected based on the type of index value you pass to the
let threeTimesTable = TimesTable(multiplier: 3)
subscript. Subscripts are not limited to a single dimension, and you can define
subscripts with multiple input parameters to suit your custom type’s needs. print("six times three is \(threeTimesTable[6])")
// Prints "six times three is 18"
Subscript Syntax In this example, a new instance of TimesTable is created to represent the three-times-
Subscripts enable you to query instances of a type by writing one or more values in table. This is indicated by passing a value of 3 to the structure’s initializer as the
square brackets after the instance name. Their syntax is similar to both instance value to use for the instance’s multiplier parameter.
method syntax and computed property syntax. You write subscript definitions with the
subscript keyword, and specify one or more input parameters and a return type, in You can query the threeTimesTable instance by calling its subscript, as shown in the
the same way as instance methods. Unlike instance methods, subscripts can be read- call to threeTimesTable[6]. This requests the sixth entry in the three-times-table,
write or read-only. This behavior is communicated by a getter and setter in the same which returns a value of 18, or 3 times 6.
way as for computed properties:
NOTE
subscript(index: Int) -> Int { An n-times-table is based on a fixed mathematical rule. It is not appropriate to set
get { threeTimesTable[someIndex] to a new value, and so the subscript for TimesTable is
defined as a read-only subscript.
// return an appropriate subscript value here
}
Subscript Usage
set(newValue) {
The exact meaning of “subscript” depends on the context in which it is used.
// perform a suitable setting action here
Subscripts are typically used as a shortcut for accessing the member elements in a
} collection, list, or sequence. You are free to implement subscripts in the most
} appropriate way for your particular class or structure’s functionality.
The type of newValue is the same as the return value of the subscript. As with
computed properties, you can choose not to specify the setter’s (newValue) For example, Swift’s Dictionary type implements a subscript to set and retrieve the
parameter. A default parameter called newValue is provided to your setter if you do not values stored in a Dictionary instance. You can set a value in a dictionary by
provide one yourself. providing a key of the dictionary’s key type within subscript brackets, and assigning a
value of the dictionary’s value type to the subscript:

95
 var numberOfLegs = ["spider": 8, "ant": 6, "cat": 4] return row >= 0 && row < rows && column >= 0 && column <
columns
numberOfLegs["bird"] = 2
}
The example above defines a variable called numberOfLegs and initializes it with a
subscript(row: Int, column: Int) -> Double {
dictionary literal containing three key-value pairs. The type of the numberOfLegs
dictionary is inferred to be [String: Int]. After creating the dictionary, this example get {
uses subscript assignment to add a String key of "bird" and an Int value of 2 to the assert(indexIsValid(row: row, column: column), "Index
dictionary. out of range")
return grid[(row * columns) + column]
For more information about Dictionary subscripting, see Accessing and Modifying a }
Dictionary. set {

NOTE assert(indexIsValid(row: row, column: column), "Index

out of range")
Swift’s Dictionary type implements its key-value subscripting as a subscript that takes grid[(row * columns) + column] = newValue
and returns an optional type. For the numberOfLegs dictionary above, the key-value
subscript takes and returns a value of type Int?, or “optional int”. The Dictionary type }
uses an optional subscript type to model the fact that not every key will have a value, }
and to give a way to delete a value for a key by assigning a nil value for that key.
}

Subscript Options Matrix provides an initializer that takes two parameters called rows and columns, and
creates an array that is large enough to store rows * columns values of type Double.
Subscripts can take any number of input parameters, and these input parameters can
Each position in the matrix is given an initial value of 0.0. To achieve this, the array’s
be of any type. Subscripts can also return any type. Subscripts can use variadic
size, and an initial cell value of 0.0, are passed to an array initializer that creates and
parameters, but they can’t use in-out parameters or provide default parameter values.
initializes a new array of the correct size. This initializer is described in more detail in
Creating an Array with a Default Value.
A class or structure can provide as many subscript implementations as it needs, and
the appropriate subscript to be used will be inferred based on the types of the value or
You can construct a new Matrix instance by passing an appropriate row and column
values that are contained within the subscript brackets at the point that the subscript is
count to its initializer:
used. This definition of multiple subscripts is known as subscript overloading.
var matrix = Matrix(rows: 2, columns: 2)
While it is most common for a subscript to take a single parameter, you can also
define a subscript with multiple parameters if it is appropriate for your type. The The preceding example creates a new Matrix instance with two rows and two
following example defines a Matrix structure, which represents a two-dimensional columns. The grid array for this Matrix instance is effectively a flattened version of
matrix of Double values. The Matrix structure’s subscript takes two integer the matrix, as read from top left to bottom right:
parameters:

struct Matrix {
let rows: Int, columns: Int
var grid: [Double]
init(rows: Int, columns: Int) {
self.rows = rows
self.columns = columns
grid = Array(repeating: 0.0, count: rows * columns)
}
func indexIsValid(row: Int, column: Int) -> Bool {

96
Values in the matrix can be set by passing row and column values into the subscript,
separated by a comma:

matrix[0, 1] = 1.5
matrix[1, 0] = 3.2

These two statements call the subscript’s setter to set a value of 1.5 in the top right
position of the matrix (where row is 0 and column is 1), and 3.2 in the bottom left
position (where row is 1 and column is 0):

The Matrix subscript’s getter and setter both contain an assertion to check that the
subscript’s row and column values are valid. To assist with these assertions, Matrix
includes a convenience method called indexIsValid(row:column:), which checks
whether the requested row and column are inside the bounds of the matrix:

func indexIsValidForRow(row: Int, column: Int) -> Bool {

return row >= 0 && row < rows && column >= 0 && column <
columns
}

An assertion is triggered if you try to access a subscript that is outside of the matrix
bounds:

let someValue = matrix[2, 2]

// this triggers an assert, because [2, 2] is outside of the matrix
bounds

97
 return "traveling at \(currentSpeed) miles per hour"
Section 13 }

Inheritance func makeNoise() {

// do nothing - an arbitrary vehicle doesn't necessarily
make a noise
}
}
Inheritance
You create a new instance of Vehicle with initializer syntax, which is written as a
A class can inherit methods, properties, and other characteristics from another class. TypeName followed by empty parentheses:
When one class inherits from another, the inheriting class is known as a subclass, and
let someVehicle = Vehicle()
the class it inherits from is known as its superclass. Inheritance is a fundamental
behavior that differentiates classes from other types in Swift. Having created a new Vehicle instance, you can access its description property to
print a human-readable description of the vehicle’s current speed:
Classes in Swift can call and access methods, properties, and subscripts belonging to
their superclass and can provide their own overriding versions of those methods, print("Vehicle: \(someVehicle.description)")
properties, and subscripts to refine or modify their behavior. Swift helps to ensure your // Vehicle: traveling at 0.0 miles per hour
overrides are correct by checking that the override definition has a matching
The Vehicle class defines common characteristics for an arbitrary vehicle, but is not
superclass definition.
much use in itself. To make it more useful, you need to refine it to describe more
specific kinds of vehicles.
Classes can also add property observers to inherited properties in order to be notified
when the value of a property changes. Property observers can be added to any
property, regardless of whether it was originally defined as a stored or computed
Subclassing
property. Subclassing is the act of basing a new class on an existing class. The subclass
inherits characteristics from the existing class, which you can then refine. You can
Defining a Base Class also add new characteristics to the subclass.

Any class that does not inherit from another class is known as a base class.
To indicate that a subclass has a superclass, write the subclass name before the
superclass name, separated by a colon:
NOTE

Swift classes do not inherit from a universal base class. Classes you define without class SomeSubclass: SomeSuperclass {
specifying a superclass automatically become base classes for you to build upon.
// subclass definition goes here
The example below defines a base class called Vehicle. This base class defines a }
stored property called currentSpeed, with a default value of 0.0 (inferring a property
The following example defines a subclass called Bicycle, with a superclass of
type of Double). The currentSpeed property’s value is used by a read-only computed
Vehicle:
String property called description to create a description of the vehicle.
class Bicycle: Vehicle {
The Vehicle base class also defines a method called makeNoise. This method does
var hasBasket = false
not actually do anything for a base Vehicle instance, but will be customized by
subclasses of Vehicle later on: }

The new Bicycle class automatically gains all of the characteristics of Vehicle, such
class Vehicle { as its currentSpeed and description properties and its makeNoise() method.
var currentSpeed = 0.0
var description: String {

98
In addition to the characteristics it inherits, the Bicycle class defines a new stored To override a characteristic that would otherwise be inherited, you prefix your
property, hasBasket, with a default value of false (inferring a type of Bool for the overriding definition with the override keyword. Doing so clarifies that you intend to
property). provide an override and have not provided a matching definition by mistake.
Overriding by accident can cause unexpected behavior, and any overrides without the
By default, any new Bicycle instance you create will not have a basket. You can set override keyword are diagnosed as an error when your code is compiled.
the hasBasket property to true for a particular Bicycle instance after that instance is
created: The override keyword also prompts the Swift compiler to check that your overriding
class’s superclass (or one of its parents) has a declaration that matches the one you
let bicycle = Bicycle() provided for the override. This check ensures that your overriding definition is correct.
bicycle.hasBasket = true
Accessing Superclass Methods, Properties, and Subscripts
You can also modify the inherited currentSpeed property of a Bicycle instance, and
query the instance’s inherited description property: When you provide a method, property, or subscript override for a subclass, it is
sometimes useful to use the existing superclass implementation as part of your
bicycle.currentSpeed = 15.0 override. For example, you can refine the behavior of that existing implementation, or
print("Bicycle: \(bicycle.description)") store a modified value in an existing inherited variable.
// Bicycle: traveling at 15.0 miles per hour
Where this is appropriate, you access the superclass version of a method, property, or
Subclasses can themselves be subclassed. The next example creates a subclass of subscript by using the super prefix:
Bicycle for a two-seater bicycle known as a “tandem”:
An overridden method named someMethod() can call the superclass
class Tandem: Bicycle { version of someMethod() by calling super.someMethod() within the
var currentNumberOfPassengers = 0 overriding method implementation.
} An overridden property called someProperty can access the superclass
Tandem inherits all of the properties and methods from Bicycle, which in turn inherits version of someProperty as super.someProperty within the overriding
all of the properties and methods from Vehicle. The Tandem subclass also adds a new getter or setter implementation.
stored property called currentNumberOfPassengers, with a default value of 0. An overridden subscript for someIndex can access the superclass version
of the same subscript as super[someIndex] from within the overriding
If you create an instance of Tandem, you can work with any of its new and inherited subscript implementation.
properties, and query the read-only description property it inherits from Vehicle:
Overriding Methods
let tandem = Tandem()
You can override an inherited instance or type method to provide a tailored or
tandem.hasBasket = true alternative implementation of the method within your subclass.
tandem.currentNumberOfPassengers = 2
tandem.currentSpeed = 22.0 The following example defines a new subclass of Vehicle called Train, which
print("Tandem: \(tandem.description)") overrides the makeNoise() method that Train inherits from Vehicle:
// Tandem: traveling at 22.0 miles per hour
class Train: Vehicle {
override func makeNoise() {
Overriding
print("Choo Choo")
A subclass can provide its own custom implementation of an instance method, type
}
method, instance property, type property, or subscript that it would otherwise inherit
from a superclass. This is known as overriding. }

If you create a new instance of Train and call its makeNoise() method, you can see
that the Train subclass version of the method is called:

99
 let train = Train() If you create an instance of the Car class and set its gear and currentSpeed
train.makeNoise() properties, you can see that its description property returns the tailored description
// Prints "Choo Choo" defined within the Car class:

Overriding Properties let car = Car()

You can override an inherited instance or type property to provide your own custom car.currentSpeed = 25.0
getter and setter for that property, or to add property observers to enable the car.gear = 3
overriding property to observe when the underlying property value changes. print("Car: \(car.description)")
// Car: traveling at 25.0 miles per hour in gear 3
Overriding Property Getters and Setters
You can provide a custom getter (and setter, if appropriate) to override any inherited Overriding Property Observers
property, regardless of whether the inherited property is implemented as a stored or You can use property overriding to add property observers to an inherited property.
computed property at source. The stored or computed nature of an inherited property This enables you to be notified when the value of an inherited property changes,
is not known by a subclass—it only knows that the inherited property has a certain regardless of how that property was originally implemented. For more information on
name and type. You must always state both the name and the type of the property you property observers, see Property Observers.
are overriding, to enable the compiler to check that your override matches a
superclass property with the same name and type. NOTE

You cannot add property observers to inherited constant stored properties or inherited
You can present an inherited read-only property as a read-write property by providing read-only computed properties. The value of these properties cannot be set, and so it is
both a getter and a setter in your subclass property override. You cannot, however, not appropriate to provide a willSet or didSet implementation as part of an override.
present an inherited read-write property as a read-only property.
Note also that you cannot provide both an overriding setter and an overriding property
observer for the same property. If you want to observe changes to a property’s value,
NOTE
and you are already providing a custom setter for that property, you can simply observe
If you provide a setter as part of a property override, you must also provide a getter for any value changes from within the custom setter.
that override. If you don’t want to modify the inherited property’s value within the
overriding getter, you can simply pass through the inherited value by returning The following example defines a new class called AutomaticCar, which is a subclass
super.someProperty from the getter, where someProperty is the name of the property of Car. The AutomaticCar class represents a car with an automatic gearbox, which
you are overriding. automatically selects an appropriate gear to use based on the current speed:
The following example defines a new class called Car, which is a subclass of Vehicle. class AutomaticCar: Car {
The Car class introduces a new stored property called gear, with a default integer
override var currentSpeed: Double {
value of 1. The Car class also overrides the description property it inherits from
Vehicle, to provide a custom description that includes the current gear: didSet {
gear = Int(currentSpeed / 10.0) + 1
class Car: Vehicle { }
var gear = 1 }
override var description: String { }
return super.description + " in gear \(gear)"
Whenever you set the currentSpeed property of an AutomaticCar instance, the
} property’s didSet observer sets the instance’s gear property to an appropriate choice
} of gear for the new speed. Specifically, the property observer chooses a gear that is
The override of the description property starts by calling super.description, which the new currentSpeed value divided by 10, rounded down to the nearest integer, plus
1. A speed of 35.0 produces a gear of 4:
returns the Vehicle class’s description property. The Car class’s version of
description then adds some extra text onto the end of this description to provide
let automatic = AutomaticCar()
information about the current gear.
automatic.currentSpeed = 35.0

100
 print("AutomaticCar: \(automatic.description)")
// AutomaticCar: traveling at 35.0 miles per hour in gear 4

Preventing Overrides
You can prevent a method, property, or subscript from being overridden by marking it
as final. Do this by writing the final modifier before the method, property, or
subscript’s introducer keyword (such as final var, final func, final class func,
and final subscript).

Any attempt to override a final method, property, or subscript in a subclass is reported

as a compile-time error. Methods, properties, or subscripts that you add to a class in
an extension can also be marked as final within the extension’s definition.

You can mark an entire class as final by writing the final modifier before the class
keyword in its class definition (final class). Any attempt to subclass a final class is
reported as a compile-time error.

101
 // perform some initialization here
Section 14 }

Initialization The example below defines a new structure called Fahrenheit to store temperatures
expressed in the Fahrenheit scale. The Fahrenheit structure has one stored property,
temperature, which is of type Double:

struct Fahrenheit {
Initialization var temperature: Double
init() {
Initialization is the process of preparing an instance of a class, structure, or
temperature = 32.0
enumeration for use. This process involves setting an initial value for each stored
property on that instance and performing any other setup or initialization that is }
required before the new instance is ready for use. }
var f = Fahrenheit()
You implement this initialization process by defining initializers, which are like special print("The default temperature is \(f.temperature)° Fahrenheit")
methods that can be called to create a new instance of a particular type. Unlike
// Prints "The default temperature is 32.0° Fahrenheit"
Objective-C initializers, Swift initializers do not return a value. Their primary role is to
ensure that new instances of a type are correctly initialized before they are used for The structure defines a single initializer, init, with no parameters, which initializes the
the first time. stored temperature with a value of 32.0 (the freezing point of water in degrees
Fahrenheit).
Instances of class types can also implement a deinitializer, which performs any
custom cleanup just before an instance of that class is deallocated. For more Default Property Values
information about deinitializers, see Deinitialization. You can set the initial value of a stored property from within an initializer, as shown
above. Alternatively, specify a default property value as part of the property’s
Setting Initial Values for Stored Properties declaration. You specify a default property value by assigning an initial value to the
Classes and structures must set all of their stored properties to an appropriate initial property when it is defined.
value by the time an instance of that class or structure is created. Stored properties
NOTE
cannot be left in an indeterminate state.
If a property always takes the same initial value, provide a default value rather than
You can set an initial value for a stored property within an initializer, or by assigning a setting a value within an initializer. The end result is the same, but the default value ties
default property value as part of the property’s definition. These actions are described the property’s initialization more closely to its declaration. It makes for shorter, clearer
initializers and enables you to infer the type of the property from its default value. The
in the following sections. default value also makes it easier for you to take advantage of default initializers and
initializer inheritance, as described later in this chapter.
NOTE
You can write the Fahrenheit structure from above in a simpler form by providing a
When you assign a default value to a stored property, or set its initial value within an
initializer, the value of that property is set directly, without calling any property
default value for its temperature property at the point that the property is declared:
observers.
struct Fahrenheit {
Initializers var temperature = 32.0
Initializers are called to create a new instance of a particular type. In its simplest form, }
an initializer is like an instance method with no parameters, written using the init
keyword: Customizing Initialization
init() {

102
You can customize the initialization process with input parameters and optional should be called. Because of this, Swift provides an automatic argument label for
property types, or by assigning constant properties during initialization, as described in every parameter in an initializer if you don’t provide one.
the following sections.
The following example defines a structure called Color, with three constant properties
Initialization Parameters called red, green, and blue. These properties store a value between 0.0 and 1.0 to
You can provide initialization parameters as part of an initializer’s definition, to define indicate the amount of red, green, and blue in the color.
the types and names of values that customize the initialization process. Initialization
parameters have the same capabilities and syntax as function and method Color provides an initializer with three appropriately named parameters of type Double
parameters. for its red, green, and blue components. Color also provides a second initializer with a
single white parameter, which is used to provide the same value for all three color
The following example defines a structure called Celsius, which stores temperatures components.
expressed in degrees Celsius. The Celsius structure implements two custom
initializers called init(fromFahrenheit:) and init(fromKelvin:), which initialize a struct Color {
new instance of the structure with a value from a different temperature scale: let red, green, blue: Double
init(red: Double, green: Double, blue: Double) {
struct Celsius {
self.red = red
var temperatureInCelsius: Double
self.green = green
init(fromFahrenheit fahrenheit: Double) {
self.blue = blue
temperatureInCelsius = (fahrenheit - 32.0) / 1.8
}
}
init(white: Double) {
init(fromKelvin kelvin: Double) {
red = white
temperatureInCelsius = kelvin - 273.15
green = white
}
blue = white
}
}
let boilingPointOfWater = Celsius(fromFahrenheit: 212.0)
}
// boilingPointOfWater.temperatureInCelsius is 100.0
Both initializers can be used to create a new Color instance, by providing named
let freezingPointOfWater = Celsius(fromKelvin: 273.15)
values for each initializer parameter:
// freezingPointOfWater.temperatureInCelsius is 0.0

The first initializer has a single initialization parameter with an argument label of let magenta = Color(red: 1.0, green: 0.0, blue: 1.0)
fromFahrenheit and a parameter name of fahrenheit. The second initializer has a let halfGray = Color(white: 0.5)
single initialization parameter with an argument label of fromKelvin and a parameter Note that it is not possible to call these initializers without using argument labels.
name of kelvin. Both initializers convert their single argument into the corresponding Argument labels must always be used in an initializer if they are defined, and omitting
Celsius value and store this value in a property called temperatureInCelsius. them is a compile-time error:

Parameter Names and Argument Labels let veryGreen = Color(0.0, 1.0, 0.0)
As with function and method parameters, initialization parameters can have both a // this reports a compile-time error - argument labels are required
parameter name for use within the initializer’s body and an argument label for use
when calling the initializer. Initializer Parameters Without Argument Labels
If you do not want to use an argument label for an initializer parameter, write an
However, initializers do not have an identifying function name before their parentheses underscore (_) instead of an explicit argument label for that parameter to override the
in the way that functions and methods do. Therefore, the names and types of an default behavior.
initializer’s parameters play a particularly important role in identifying which initializer

103
Here’s an expanded version of the Celsius example from earlier, with an additional }
initializer to create a new Celsius instance from a Double value that is already in the let cheeseQuestion = SurveyQuestion(text: "Do you like cheese?")
Celsius scale: cheeseQuestion.ask()
// Prints "Do you like cheese?"
struct Celsius {
cheeseQuestion.response = "Yes, I do like cheese."
var temperatureInCelsius: Double
init(fromFahrenheit fahrenheit: Double) { The response to a survey question cannot be known until it is asked, and so the
response property is declared with a type of String?, or “optional String”. It is
temperatureInCelsius = (fahrenheit - 32.0) / 1.8
automatically assigned a default value of nil, meaning “no string yet”, when a new
} instance of SurveyQuestion is initialized.
init(fromKelvin kelvin: Double) {
temperatureInCelsius = kelvin - 273.15 Assigning Constant Properties During Initialization
} You can assign a value to a constant property at any point during initialization, as long
init(_ celsius: Double) { as it is set to a definite value by the time initialization finishes. Once a constant
property is assigned a value, it can’t be further modified.
temperatureInCelsius = celsius
} NOTE
}
For class instances, a constant property can be modified during initialization only by the
let bodyTemperature = Celsius(37.0) class that introduces it. It cannot be modified by a subclass.
// bodyTemperature.temperatureInCelsius is 37.0
You can revise the SurveyQuestion example from above to use a constant property
The initializer call Celsius(37.0) is clear in its intent without the need for an argument rather than a variable property for the text property of the question, to indicate that
label. It is therefore appropriate to write this initializer as init(_ celsius: Double) so the question does not change once an instance of SurveyQuestion is created. Even
that it can be called by providing an unnamed Double value. though the text property is now a constant, it can still be set within the class’s
initializer:
Optional Property Types
class SurveyQuestion {
If your custom type has a stored property that is logically allowed to have “no value”—
perhaps because its value cannot be set during initialization, or because it is allowed let text: String
to have “no value” at some later point—declare the property with an optional type. var response: String?
Properties of optional type are automatically initialized with a value of nil, indicating init(text: String) {
that the property is deliberately intended to have “no value yet” during initialization.
self.text = text
}
The following example defines a class called SurveyQuestion, with an optional String
property called response: func ask() {
print(text)
class SurveyQuestion {
}
var text: String
}
var response: String?
let beetsQuestion = SurveyQuestion(text: "How about beets?")
init(text: String) {
beetsQuestion.ask()
self.text = text
// Prints "How about beets?"
}
beetsQuestion.response = "I also like beets. (But not with cheese.)"
func ask() {
print(text) Default Initializers
}

104
Swift provides a default initializer for any structure or class that provides default values Initializer Delegation for Value Types
for all of its properties and does not provide at least one initializer itself. The default
Initializers can call other initializers to perform part of an instance’s initialization. This
initializer simply creates a new instance with all of its properties set to their default
process, known as initializer delegation, avoids duplicating code across multiple
values.
initializers.

This example defines a class called ShoppingListItem, which encapsulates the name,
The rules for how initializer delegation works, and for what forms of delegation are
quantity, and purchase state of an item in a shopping list:
allowed, are different for value types and class types. Value types (structures and
enumerations) do not support inheritance, and so their initializer delegation process is
class ShoppingListItem {
relatively simple, because they can only delegate to another initializer that they
var name: String? provide themselves. Classes, however, can inherit from other classes, as described in
var quantity = 1 Inheritance. This means that classes have additional responsibilities for ensuring that
var purchased = false all stored properties they inherit are assigned a suitable value during initialization.
} These responsibilities are described in Class Inheritance and Initialization below.
var item = ShoppingListItem()
For value types, you use self.init to refer to other initializers from the same value
Because all properties of the ShoppingListItem class have default values, and type when writing your own custom initializers. You can call self.init only from
because it is a base class with no superclass, ShoppingListItem automatically gains within an initializer.
a default initializer implementation that creates a new instance with all of its properties
set to their default values. (The name property is an optional String property, and so it Note that if you define a custom initializer for a value type, you will no longer have
automatically receives a default value of nil, even though this value is not written in access to the default initializer (or the memberwise initializer, if it is a structure) for that
the code.) The example above uses the default initializer for the ShoppingListItem type. This constraint prevents a situation in which additional essential setup provided
class to create a new instance of the class with initializer syntax, written as in a more complex initializer is accidentally circumvented by someone using one of the
ShoppingListItem(), and assigns this new instance to a variable called item. automatic initializers.

Memberwise Initializers for Structure Types NOTE

Structure types automatically receive a memberwise initializer if they do not define any If you want your custom value type to be initializable with the default initializer and
of their own custom initializers. Unlike a default initializer, the structure receives a memberwise initializer, and also with your own custom initializers, write your custom
memberwise initializer even if it has stored properties that do not have default values. initializers in an extension rather than as part of the value type’s original
implementation. For more information, see Extensions.
The memberwise initializer is a shorthand way to initialize the member properties of The following example defines a custom Rect structure to represent a geometric
new structure instances. Initial values for the properties of the new instance can be rectangle. The example requires two supporting structures called Size and Point,
passed to the memberwise initializer by name. both of which provide default values of 0.0 for all of their properties:

The example below defines a structure called Size with two properties called width struct Size {
and height. Both properties are inferred to be of type Double by assigning a default var width = 0.0, height = 0.0
value of 0.0.
}

The Size structure automatically receives an init(width:height:) memberwise struct Point {

initializer, which you can use to initialize a new Size instance: var x = 0.0, y = 0.0
}
struct Size {
You can initialize the Rect structure below in one of three ways—by using its default
var width = 0.0, height = 0.0
zero-initialized origin and size property values, by providing a specific origin point
} and size, or by providing a specific center point and size. These initialization options
let twoByTwo = Size(width: 2.0, height: 2.0)

105
are represented by three custom initializers that are part of the Rect structure’s // centerRect's origin is (2.5, 2.5) and its size is (3.0, 3.0)
definition: The init(center:size:) initializer could have assigned the new values of origin and
size to the appropriate properties itself. However, it is more convenient (and clearer in
struct Rect { intent) for the init(center:size:) initializer to take advantage of an existing initializer
var origin = Point() that already provides exactly that functionality.
var size = Size()
NOTE
init() {}
init(origin: Point, size: Size) { For an alternative way to write this example without defining the init() and
init(origin:size:) initializers yourself, see Extensions.
self.origin = origin
self.size = size
Class Inheritance and Initialization
}
All of a class’s stored properties—including any properties the class inherits from its
init(center: Point, size: Size) { superclass—must be assigned an initial value during initialization.
let originX = center.x - (size.width / 2)
let originY = center.y - (size.height / 2) Swift defines two kinds of initializers for class types to help ensure all stored
self.init(origin: Point(x: originX, y: originY), size: size)
properties receive an initial value. These are known as designated initializers and
convenience initializers.
}
} Designated Initializers and Convenience Initializers
The first Rect initializer, init(), is functionally the same as the default initializer that Designated initializers are the primary initializers for a class. A designated initializer
the structure would have received if it did not have its own custom initializers. This fully initializes all properties introduced by that class and calls an appropriate
initializer has an empty body, represented by an empty pair of curly braces {}. Calling superclass initializer to continue the initialization process up the superclass chain.
this initializer returns a Rect instance whose origin and size properties are both
initialized with the default values of Point(x: 0.0, y: 0.0) and Size(width: 0.0, Classes tend to have very few designated initializers, and it is quite common for a
height: 0.0) from their property definitions: class to have only one. Designated initializers are “funnel” points through which
initialization takes place, and through which the initialization process continues up the
let basicRect = Rect() superclass chain.
// basicRect's origin is (0.0, 0.0) and its size is (0.0, 0.0)

The second Rect initializer, init(origin:size:), is functionally the same as the Every class must have at least one designated initializer. In some cases, this
memberwise initializer that the structure would have received if it did not have its own requirement is satisfied by inheriting one or more designated initializers from a
custom initializers. This initializer simply assigns the origin and size argument superclass, as described in Automatic Initializer Inheritance below.
values to the appropriate stored properties:
Convenience initializers are secondary, supporting initializers for a class. You can
let originRect = Rect(origin: Point(x: 2.0, y: 2.0), define a convenience initializer to call a designated initializer from the same class as
size: Size(width: 5.0, height: 5.0)) the convenience initializer with some of the designated initializer’s parameters set to
default values. You can also define a convenience initializer to create an instance of
// originRect's origin is (2.0, 2.0) and its size is (5.0, 5.0)
that class for a specific use case or input value type.
The third Rect initializer, init(center:size:), is slightly more complex. It starts by
calculating an appropriate origin point based on a center point and a size value. It You do not have to provide convenience initializers if your class does not require
then calls (or delegates) to the init(origin:size:) initializer, which stores the new them. Create convenience initializers whenever a shortcut to a common initialization
origin and size values in the appropriate properties: pattern will save time or make initialization of the class clearer in intent.

let centerRect = Rect(center: Point(x: 4.0, y: 4.0),

Syntax for Designated and Convenience Initializers
size: Size(width: 3.0, height: 3.0))

106
Designated initializers for classes are written in the same way as simple initializers for Here, the superclass has a single designated initializer and two convenience
value types: initializers. One convenience initializer calls another convenience initializer, which in
turn calls the single designated initializer. This satisfies rules 2 and 3 from above. The
init(parameters) { superclass does not itself have a further superclass, and so rule 1 does not apply.
statements
}
The subclass in this figure has two designated initializers and one convenience
Convenience initializers are written in the same style, but with the convenience
initializer. The convenience initializer must call one of the two designated initializers,
modifier placed before the init keyword, separated by a space:
because it can only call another initializer from the same class. This satisfies rules 2
and 3 from above. Both designated initializers must call the single designated
convenience init(parameters) {
statements initializer from the superclass, to satisfy rule 1 from above.
}
NOTE
Initializer Delegation for Class Types
These rules don’t affect how users of your classes create instances of each class. Any
To simplify the relationships between designated and convenience initializers, Swift initializer in the diagram above can be used to create a fully-initialized instance of the
applies the following three rules for delegation calls between initializers: class they belong to. The rules only affect how you write the implementation of the
class’s initializers.
Rule 1 The figure below shows a more complex class hierarchy for four classes. It illustrates
A designated initializer must call a designated initializer from its immediate how the designated initializers in this hierarchy act as “funnel” points for class
superclass. initialization, simplifying the interrelationships among classes in the chain:

Rule 2
A convenience initializer must call another initializer from the same class.

Rule 3
A convenience initializer must ultimately call a designated initializer.

A simple way to remember this is:

Designated initializers must always delegate up.

Convenience initializers must always delegate across.
These rules are illustrated in the figure below:

Two-Phase Initialization
Class initialization in Swift is a two-phase process. In the first phase, each stored
property is assigned an initial value by the class that introduced it. Once the initial
state for every stored property has been determined, the second phase begins, and

107
each class is given the opportunity to customize its stored properties further before the The class instance is not fully valid until the first phase ends. Properties can only be
new instance is considered ready for use. accessed, and methods can only be called, once the class instance is known to be
valid at the end of the first phase.
The use of a two-phase initialization process makes initialization safe, while still giving
complete flexibility to each class in a class hierarchy. Two-phase initialization prevents Here’s how two-phase initialization plays out, based on the four safety checks above:
property values from being accessed before they are initialized, and prevents property
values from being set to a different value by another initializer unexpectedly. Phase 1

NOTE A designated or convenience initializer is called on a class.

Swift’s two-phase initialization process is similar to initialization in Objective-C. The Memory for a new instance of that class is allocated. The memory is not yet
main difference is that during phase 1, Objective-C assigns zero or null values (such as
initialized.
0 or nil) to every property. Swift’s initialization flow is more flexible in that it lets you set
custom initial values, and can cope with types for which 0 or nil is not a valid default A designated initializer for that class confirms that all stored properties
value. introduced by that class have a value. The memory for these stored
Swift’s compiler performs four helpful safety-checks to make sure that two-phase properties is now initialized.
initialization is completed without error: The designated initializer hands off to a superclass initializer to perform the
same task for its own stored properties.
Safety check 1
This continues up the class inheritance chain until the top of the chain is
A designated initializer must ensure that all of the properties introduced by its reached.
class are initialized before it delegates up to a superclass initializer.
Once the top of the chain is reached, and the final class in the chain has
As mentioned above, the memory for an object is only considered fully initialized once ensured that all of its stored properties have a value, the instance’s
the initial state of all of its stored properties is known. In order for this rule to be memory is considered to be fully initialized, and phase 1 is complete.
satisfied, a designated initializer must make sure that all of its own properties are Phase 2
initialized before it hands off up the chain.
Working back down from the top of the chain, each designated initializer in
Safety check 2 the chain has the option to customize the instance further. Initializers are
A designated initializer must delegate up to a superclass initializer before now able to access self and can modify its properties, call its instance
assigning a value to an inherited property. If it doesn’t, the new value the methods, and so on.
designated initializer assigns will be overwritten by the superclass as part of Finally, any convenience initializers in the chain have the option to
its own initialization. customize the instance and to work with self.

Safety check 3 Here’s how phase 1 looks for an initialization call for a hypothetical subclass and
superclass:
A convenience initializer must delegate to another initializer before assigning a
value to any property (including properties defined by the same class). If it
doesn’t, the new value the convenience initializer assigns will be overwritten
by its own class’s designated initializer.

Safety check 4
An initializer cannot call any instance methods, read the values of any
instance properties, or refer to self as a value until after the first phase of
initialization is complete.

108
In this example, initialization begins with a call to a convenience initializer on the Superclass initializers are inherited in certain circumstances, but only when it is safe
subclass. This convenience initializer cannot yet modify any properties. It delegates and appropriate to do so. For more information, see Automatic Initializer Inheritance
below.
across to a designated initializer from the same class.
If you want a custom subclass to present one or more of the same initializers as its
The designated initializer makes sure that all of the subclass’s properties have a superclass, you can provide a custom implementation of those initializers within the
value, as per safety check 1. It then calls a designated initializer on its superclass to subclass.
continue the initialization up the chain.
When you write a subclass initializer that matches a superclass designated initializer,
The superclass’s designated initializer makes sure that all of the superclass properties you are effectively providing an override of that designated initializer. Therefore, you
have a value. There are no further superclasses to initialize, and so no further must write the override modifier before the subclass’s initializer definition. This is true
delegation is needed. even if you are overriding an automatically provided default initializer, as described in
Default Initializers.
As soon as all properties of the superclass have an initial value, its memory is
considered fully initialized, and Phase 1 is complete. As with an overridden property, method or subscript, the presence of the override
modifier prompts Swift to check that the superclass has a matching designated
Here’s how phase 2 looks for the same initialization call: initializer to be overridden, and validates that the parameters for your overriding
initializer have been specified as intended.

NOTE

You always write the override modifier when overriding a superclass designated
initializer, even if your subclass’s implementation of the initializer is a convenience
initializer.

Conversely, if you write a subclass initializer that matches a superclass convenience

initializer, that superclass convenience initializer can never be called directly by your
subclass, as per the rules described above in Initializer Delegation for Class Types.
Therefore, your subclass is not (strictly speaking) providing an override of the
superclass initializer. As a result, you do not write the override modifier when
providing a matching implementation of a superclass convenience initializer.
The superclass’s designated initializer now has an opportunity to customize the
instance further (although it does not have to). The example below defines a base class called Vehicle. This base class declares a
stored property called numberOfWheels, with a default Int value of 0. The
Once the superclass’s designated initializer is finished, the subclass’s designated numberOfWheels property is used by a computed property called description to
initializer can perform additional customization (although again, it does not have to). create a String description of the vehicle’s characteristics:

Finally, once the subclass’s designated initializer is finished, the convenience initializer class Vehicle {
that was originally called can perform additional customization. var numberOfWheels = 0
var description: String {
Initializer Inheritance and Overriding
return "\(numberOfWheels) wheel(s)"
Unlike subclasses in Objective-C, Swift subclasses do not inherit their superclass
}
initializers by default. Swift’s approach prevents a situation in which a simple initializer
from a superclass is inherited by a more specialized subclass and is used to create a }
new instance of the subclass that is not fully or correctly initialized. The Vehicle class provides a default value for its only stored property, and does not
provide any custom initializers itself. As a result, it automatically receives a default
NOTE initializer, as described in Default Initializers. The default initializer (when available) is

109
always a designated initializer for a class, and can be used to create a new Vehicle Assuming that you provide default values for any new properties you introduce in a
instance with a numberOfWheels of 0: subclass, the following two rules apply:

let vehicle = Vehicle() Rule 1

print("Vehicle: \(vehicle.description)") If your subclass doesn’t define any designated initializers, it automatically
// Vehicle: 0 wheel(s) inherits all of its superclass designated initializers.
The next example defines a subclass of Vehicle called Bicycle:
Rule 2
class Bicycle: Vehicle { If your subclass provides an implementation of all of its superclass designated
override init() { initializers—either by inheriting them as per rule 1, or by providing a custom
super.init()
implementation as part of its definition—then it automatically inherits all of the
superclass convenience initializers.
numberOfWheels = 2
} These rules apply even if your subclass adds further convenience initializers.
}
NOTE
The Bicycle subclass defines a custom designated initializer, init(). This designated
initializer matches a designated initializer from the superclass of Bicycle, and so the A subclass can implement a superclass designated initializer as a subclass
convenience initializer as part of satisfying rule 2.
Bicycle version of this initializer is marked with the override modifier.

Designated and Convenience Initializers in Action

The init() initializer for Bicycle starts by calling super.init(), which calls the
The following example shows designated initializers, convenience initializers, and
default initializer for the Bicycle class’s superclass, Vehicle. This ensures that the
automatic initializer inheritance in action. This example defines a hierarchy of three
numberOfWheels inherited property is initialized by Vehicle before Bicycle has the
classes called Food, RecipeIngredient, and ShoppingListItem, and demonstrates
opportunity to modify the property. After calling super.init(), the original value of
how their initializers interact.
numberOfWheels is replaced with a new value of 2.

The base class in the hierarchy is called Food, which is a simple class to encapsulate
If you create an instance of Bicycle, you can call its inherited description computed
the name of a foodstuff. The Food class introduces a single String property called
property to see how its numberOfWheels property has been updated:
name and provides two initializers for creating Food instances:
let bicycle = Bicycle()
class Food {
print("Bicycle: \(bicycle.description)")
var name: String
// Bicycle: 2 wheel(s)
init(name: String) {
NOTE
self.name = name
Subclasses can modify inherited variable properties during initialization, but can not
}
modify inherited constant properties.
convenience init() {
Automatic Initializer Inheritance self.init(name: "[Unnamed]")
As mentioned above, subclasses do not inherit their superclass initializers by default. }
However, superclass initializers are automatically inherited if certain conditions are }
met. In practice, this means that you do not need to write initializer overrides in many
common scenarios, and can inherit your superclass initializers with minimal effort The figure below shows the initializer chain for the Food class:
whenever it is safe to do so.

110
Classes do not have a default memberwise initializer, and so the Food class provides
a designated initializer that takes a single argument called name. This initializer can be
used to create a new Food instance with a specific name:

let namedMeat = Food(name: "Bacon")

// namedMeat's name is "Bacon"
The RecipeIngredient class has a single designated initializer, init(name: String,
The init(name: String) initializer from the Food class is provided as a designated quantity: Int), which can be used to populate all of the properties of a new
initializer, because it ensures that all stored properties of a new Food instance are fully RecipeIngredient instance. This initializer starts by assigning the passed quantity
initialized. The Food class does not have a superclass, and so the init(name: argument to the quantity property, which is the only new property introduced by
String) initializer does not need to call super.init() to complete its initialization. RecipeIngredient. After doing so, the initializer delegates up to the init(name:
String) initializer of the Food class. This process satisfies safety check 1 from Two-
The Food class also provides a convenience initializer, init(), with no arguments. Phase Initialization above.
The init() initializer provides a default placeholder name for a new food by
delegating across to the Food class’s init(name: String) with a name value of RecipeIngredient also defines a convenience initializer, init(name: String), which
[Unnamed]: is used to create a RecipeIngredient instance by name alone. This convenience
initializer assumes a quantity of 1 for any RecipeIngredient instance that is created
let mysteryMeat = Food() without an explicit quantity. The definition of this convenience initializer makes
// mysteryMeat's name is "[Unnamed]" RecipeIngredient instances quicker and more convenient to create, and avoids code
The second class in the hierarchy is a subclass of Food called RecipeIngredient. The duplication when creating several single-quantity RecipeIngredient instances. This
RecipeIngredient class models an ingredient in a cooking recipe. It introduces an Int convenience initializer simply delegates across to the class’s designated initializer,
property called quantity (in addition to the name property it inherits from Food) and passing in a quantity value of 1.
defines two initializers for creating RecipeIngredient instances:
The init(name: String) convenience initializer provided by RecipeIngredient takes
class RecipeIngredient: Food { the same parameters as the init(name: String) designated initializer from Food.
var quantity: Int Because this convenience initializer overrides a designated initializer from its
superclass, it must be marked with the override modifier (as described in Initializer
init(name: String, quantity: Int) {
Inheritance and Overriding).
self.quantity = quantity
super.init(name: name) Even though RecipeIngredient provides the init(name: String) initializer as a
} convenience initializer, RecipeIngredient has nonetheless provided an
override convenience init(name: String) { implementation of all of its superclass’s designated initializers. Therefore,
RecipeIngredient automatically inherits all of its superclass’s convenience initializers
self.init(name: name, quantity: 1)
too.
}
} In this example, the superclass for RecipeIngredient is Food, which has a single
The figure below shows the initializer chain for the RecipeIngredient class: convenience initializer called init(). This initializer is therefore inherited by
RecipeIngredient. The inherited version of init() functions in exactly the same way

111
as the Food version, except that it delegates to the RecipeIngredient version of
init(name: String) rather than the Food version.

All three of these initializers can be used to create new RecipeIngredient instances:

let oneMysteryItem = RecipeIngredient()

let oneBacon = RecipeIngredient(name: "Bacon")
let sixEggs = RecipeIngredient(name: "Eggs", quantity: 6)

The third and final class in the hierarchy is a subclass of RecipeIngredient called
ShoppingListItem. The ShoppingListItem class models a recipe ingredient as it
appears in a shopping list.

Every item in the shopping list starts out as “unpurchased”. To represent this fact,
ShoppingListItem introduces a Boolean property called purchased, with a default
value of false. ShoppingListItem also adds a computed description property, which
provides a textual description of a ShoppingListItem instance:

class ShoppingListItem: RecipeIngredient {

You can use all three of the inherited initializers to create a new ShoppingListItem
var purchased = false
instance:
var description: String {
var output = "\(quantity) x \(name)" var breakfastList = [
output += purchased ? " ✔" : " ✘" ShoppingListItem(),
return output ShoppingListItem(name: "Bacon"),
} ShoppingListItem(name: "Eggs", quantity: 6),
} ]
NOTE breakfastList[0].name = "Orange juice"

ShoppingListItem does not define an initializer to provide an initial value for breakfastList[0].purchased = true
purchased, because items in a shopping list (as modeled here) always start out for item in breakfastList {
unpurchased.
print(item.description)
Because it provides a default value for all of the properties it introduces and does not }
define any initializers itself, ShoppingListItem automatically inherits all of the
// 1 x Orange juice ✔
designated and convenience initializers from its superclass.
// 1 x Bacon ✘

The figure below shows the overall initializer chain for all three classes: // 6 x Eggs ✘

Here, a new array called breakfastList is created from an array literal containing
three new ShoppingListItem instances. The type of the array is inferred to be
[ShoppingListItem]. After the array is created, the name of the ShoppingListItem at
the start of the array is changed from "[Unnamed]" to "Orange juice" and it is marked
as having been purchased. Printing the description of each item in the array shows
that their default states have been set as expected.

Failable Initializers

112
It is sometimes useful to define a class, structure, or enumeration for which print("An animal was initialized with a species of \
(giraffe.species)")
initialization can fail. This failure might be triggered by invalid initialization parameter
values, the absence of a required external resource, or some other condition that }
prevents initialization from succeeding. // Prints "An animal was initialized with a species of Giraffe"

If you pass an empty string value to the failable initializer’s species parameter, the
To cope with initialization conditions that can fail, define one or more failable initializer triggers an initialization failure:
initializers as part of a class, structure, or enumeration definition. You write a failable
initializer by placing a question mark after the init keyword (init?). let anonymousCreature = Animal(species: "")
// anonymousCreature is of type Animal?, not Animal
NOTE

You cannot define a failable and a nonfailable initializer with the same parameter types
and names. if anonymousCreature == nil {
print("The anonymous creature could not be initialized")
A failable initializer creates an optional value of the type it initializes. You write return
}
nil within a failable initializer to indicate a point at which initialization failure can be
triggered. // Prints "The anonymous creature could not be initialized"
NOTE
NOTE
Checking for an empty string value (such as "" rather than "Giraffe") is not the same
Strictly speaking, initializers do not return a value. Rather, their role is to ensure that as checking for nil to indicate the absence of an optional String value. In the
self is fully and correctly initialized by the time that initialization ends. Although you example above, an empty string ("") is a valid, nonoptional String. However, it is not
write return nil to trigger an initialization failure, you do not use the return keyword appropriate for an animal to have an empty string as the value of its species property.
to indicate initialization success. To model this restriction, the failable initializer triggers an initialization failure if an empty
string is found.
The example below defines a structure called Animal, with a constant String property
called species. The Animal structure also defines a failable initializer with a single Failable Initializers for Enumerations
parameter called species. This initializer checks if the species value passed to the You can use a failable initializer to select an appropriate enumeration case based on
initializer is an empty string. If an empty string is found, an initialization failure is one or more parameters. The initializer can then fail if the provided parameters do not
triggered. Otherwise, the species property’s value is set, and initialization succeeds: match an appropriate enumeration case.
struct Animal {
The example below defines an enumeration called TemperatureUnit, with three
let species: String
possible states (kelvin, celsius, and fahrenheit). A failable initializer is used to find
init?(species: String) { an appropriate enumeration case for a Character value representing a temperature
if species.isEmpty { return nil } symbol:
self.species = species
enum TemperatureUnit {
}
case kelvin, celsius, fahrenheit
}
init?(symbol: Character) {
You can use this failable initializer to try to initialize a new Animal instance and to
switch symbol {
check if initialization succeeded:
case "K":
let someCreature = Animal(species: "Giraffe") self = .kelvin
// someCreature is of type Animal?, not Animal case "C":
self = .celsius
if let giraffe = someCreature { case "F":
self = .fahrenheit

113
 default: }
return nil // Prints "This is a defined temperature unit, so initialization
succeeded."
}
}
let unknownUnit = TemperatureUnit(rawValue: "X")
}
if unknownUnit == nil {
You can use this failable initializer to choose an appropriate enumeration case for the
print("This is not a defined temperature unit, so initialization
three possible states and to cause initialization to fail if the parameter does not match failed.")
one of these states:
}
let fahrenheitUnit = TemperatureUnit(symbol: "F") // Prints "This is not a defined temperature unit, so initialization
failed."
if fahrenheitUnit != nil {
print("This is a defined temperature unit, so initialization Propagation of Initialization Failure
succeeded.")
A failable initializer of a class, structure, or enumeration can delegate across to
}
another failable initializer from the same class, structure, or enumeration. Similarly, a
// Prints "This is a defined temperature unit, so initialization subclass failable initializer can delegate up to a superclass failable initializer.
succeeded."

In either case, if you delegate to another initializer that causes initialization to fail, the
let unknownUnit = TemperatureUnit(symbol: "X") entire initialization process fails immediately, and no further initialization code is
if unknownUnit == nil { executed.
print("This is not a defined temperature unit, so
NOTE
initialization failed.")
} A failable initializer can also delegate to a nonfailable initializer. Use this approach if
you need to add a potential failure state to an existing initialization process that does
// Prints "This is not a defined temperature unit, so initialization
not otherwise fail.
failed."
The example below defines a subclass of Product called CartItem. The CartItem
Failable Initializers for Enumerations with Raw Values class models an item in an online shopping cart. CartItem introduces a stored
Enumerations with raw values automatically receive a failable initializer, init? constant property called quantity and ensures that this property always has a value
(rawValue:), that takes a parameter called rawValue of the appropriate raw-value of at least 1:
type and selects a matching enumeration case if one is found, or triggers an
initialization failure if no matching value exists. class Product {
let name: String
You can rewrite the TemperatureUnit example from above to use raw values of type init?(name: String) {
Character and to take advantage of the init?(rawValue:) initializer:
if name.isEmpty { return nil }
enum TemperatureUnit: Character { self.name = name
case kelvin = "K", celsius = "C", fahrenheit = "F" }
} }

let fahrenheitUnit = TemperatureUnit(rawValue: "F") class CartItem: Product {

if fahrenheitUnit != nil { let quantity: Int
print("This is a defined temperature unit, so initialization init?(name: String, quantity: Int) {
succeeded.") if quantity < 1 { return nil }

114
 self.quantity = quantity subclass nonfailable initializer. This enables you to define a subclass for which
super.init(name: name) initialization cannot fail, even though initialization of the superclass is allowed to fail.
}
}
Note that if you override a failable superclass initializer with a nonfailable subclass
initializer, the only way to delegate up to the superclass initializer is to force-unwrap
The failable initializer for CartItem starts by validating that it has received a quantity the result of the failable superclass initializer.
value of 1 or more. If the quantity is invalid, the entire initialization process fails
immediately and no further initialization code is executed. Likewise, the failable NOTE
initializer for Product checks the name value, and the initializer process fails
You can override a failable initializer with a nonfailable initializer but not the other way
immediately if name is the empty string. around.

If you create a CartItem instance with a nonempty name and a quantity of 1 or more, The example below defines a class called Document. This class models a document
initialization succeeds: that can be initialized with a name property that is either a nonempty string value or
nil, but cannot be an empty string:
if let twoSocks = CartItem(name: "sock", quantity: 2) {
class Document {
print("Item: \(twoSocks.name), quantity: \(twoSocks.quantity)")
var name: String?
}
// this initializer creates a document with a nil name value
// Prints "Item: sock, quantity: 2"
init() {}
If you try to create a CartItem instance with a quantity value of 0, the CartItem
initializer causes initialization to fail: // this initializer creates a document with a nonempty name
value

if let zeroShirts = CartItem(name: "shirt", quantity: 0) { init?(name: String) {

print("Item: \(zeroShirts.name), quantity: \ if name.isEmpty { return nil }

(zeroShirts.quantity)") self.name = name
} else { }
print("Unable to initialize zero shirts") }
}
The next example defines a subclass of Document called
// Prints "Unable to initialize zero shirts" AutomaticallyNamedDocument. The AutomaticallyNamedDocument subclass overrides
Similarly, if you try to create a CartItem instance with an empty name value, the both of the designated initializers introduced by Document. These overrides ensure that
superclass Product initializer causes initialization to fail: an AutomaticallyNamedDocument instance has an initial name value of "[Untitled]" if
the instance is initialized without a name, or if an empty string is passed to the
if let oneUnnamed = CartItem(name: "", quantity: 1) { init(name:) initializer:
print("Item: \(oneUnnamed.name), quantity: \
(oneUnnamed.quantity)") class AutomaticallyNamedDocument: Document {

} else { override init() {

print("Unable to initialize one unnamed product") super.init()

} self.name = "[Untitled]"

// Prints "Unable to initialize one unnamed product" }

override init(name: String) {
Overriding a Failable Initializer super.init()
You can override a superclass failable initializer in a subclass, just like any other if name.isEmpty {
initializer. Alternatively, you can override a superclass failable initializer with a
self.name = "[Untitled]"

115
 } else { required init() {
self.name = name // initializer implementation goes here
} }
} }
} You must also write the required modifier before every subclass implementation of a
The AutomaticallyNamedDocument overrides its superclass’s failable init?(name:) required initializer, to indicate that the initializer requirement applies to further
initializer with a nonfailable init(name:) initializer. Because subclasses in the chain. You do not write the override modifier when overriding a
AutomaticallyNamedDocument copes with the empty string case in a different way than required designated initializer:
its superclass, its initializer does not need to fail, and so it provides a nonfailable
version of the initializer instead. class SomeSubclass: SomeClass {
required init() {
You can use forced unwrapping in an initializer to call a failable initializer from the // subclass implementation of the required initializer goes
superclass as part of the implementation of a subclass’s nonfailable initializer. For here
example, the UntitledDocument subclass below is always named "[Untitled]", and }
it uses the failable init(name:) initializer from its superclass during initialization. }
NOTE
class UntitledDocument: Document {
override init() { You do not have to provide an explicit implementation of a required initializer if you can
satisfy the requirement with an inherited initializer.
super.init(name: "[Untitled]")!
}
Setting a Default Property Value with a Closure or Function
}
If a stored property’s default value requires some customization or setup, you can use
In this case, if the init(name:) initializer of the superclass were ever called with an a closure or global function to provide a customized default value for that property.
empty string as the name, the forced unwrapping operation would result in a runtime Whenever a new instance of the type that the property belongs to is initialized, the
error. However, because it’s called with a string constant, you can see that the closure or function is called, and its return value is assigned as the property’s default
initializer won’t fail, so no runtime error can occur in this case. value.

The init! Failable Initializer These kinds of closures or functions typically create a temporary value of the same
You typically define a failable initializer that creates an optional instance of the type as the property, tailor that value to represent the desired initial state, and then
appropriate type by placing a question mark after the init keyword (init?). return that temporary value to be used as the property’s default value.
Alternatively, you can define a failable initializer that creates an implicitly unwrapped
optional instance of the appropriate type. Do this by placing an exclamation mark after Here’s a skeleton outline of how a closure can be used to provide a default property
the init keyword (init!) instead of a question mark. value:

You can delegate from init? to init! and vice versa, and you can override init? class SomeClass {
with init! and vice versa. You can also delegate from init to init!, although doing let someProperty: SomeType = {
so will trigger an assertion if the init! initializer causes initialization to fail. // create a default value for someProperty inside this
closure
Required Initializers // someValue must be of the same type as SomeType

Write the required modifier before the definition of a class initializer to indicate that return someValue
every subclass of the class must implement that initializer: }()
}
class SomeClass {

116
Note that the closure’s end curly brace is followed by an empty pair of parentheses. var isBlack = false
This tells Swift to execute the closure immediately. If you omit these parentheses, you for i in 1...8 {
are trying to assign the closure itself to the property, and not the return value of the for j in 1...8 {
closure.
temporaryBoard.append(isBlack)

NOTE isBlack = !isBlack

}
If you use a closure to initialize a property, remember that the rest of the instance has
not yet been initialized at the point that the closure is executed. This means that you isBlack = !isBlack
cannot access any other property values from within your closure, even if those }
properties have default values. You also cannot use the implicit self property, or call
any of the instance’s methods. return temporaryBoard
}()
The example below defines a structure called Chessboard, which models a board for
the game of chess. Chess is played on an 8 x 8 board, with alternating black and func squareIsBlackAt(row: Int, column: Int) -> Bool {
white squares. return boardColors[(row * 8) + column]
}
}

Whenever a new Chessboard instance is created, the closure is executed, and the
default value of boardColors is calculated and returned. The closure in the example
above calculates and sets the appropriate color for each square on the board in a
temporary array called temporaryBoard, and returns this temporary array as the
closure’s return value once its setup is complete. The returned array value is stored in
boardColors and can be queried with the squareIsBlackAtRow utility function:

let board = Chessboard()

print(board.squareIsBlackAt(row: 0, column: 1))
// Prints "true"
print(board.squareIsBlackAt(row: 7, column: 7))
// Prints "false"

To represent this game board, the Chessboard structure has a single property called
boardColors, which is an array of 64 Bool values. A value of true in the array
represents a black square and a value of false represents a white square. The first
item in the array represents the top left square on the board and the last item in the
array represents the bottom right square on the board.

The boardColors array is initialized with a closure to set up its color values:

struct Chessboard {
let boardColors: [Bool] = {
var temporaryBoard = [Bool]()

117
 Section 15 which can never have more than 10,000 coins in circulation. There can only ever be
one Bank in the game, and so the Bank is implemented as a class with type properties

Deinitialization and methods to store and manage its current state:

class Bank {
static var coinsInBank = 10_000
static func distribute(coins numberOfCoinsRequested: Int) ->
Deinitialization Int {
let numberOfCoinsToVend = min(numberOfCoinsRequested,
A deinitializer is called immediately before a class instance is deallocated. You write coinsInBank)
deinitializers with the deinit keyword, similar to how initializers are written with the coinsInBank -= numberOfCoinsToVend
init keyword. Deinitializers are only available on class types. return numberOfCoinsToVend
}
How Deinitialization Works static func receive(coins: Int) {
Swift automatically deallocates your instances when they are no longer needed, to coinsInBank += coins
free up resources. Swift handles the memory management of instances through
}
automatic reference counting (ARC), as described in Automatic Reference Counting.
Typically you don’t need to perform manual cleanup when your instances are }
deallocated. However, when you are working with your own resources, you might Bank keeps track of the current number of coins it holds with its coinsInBank property.
need to perform some additional cleanup yourself. For example, if you create a It also offers two methods—distribute(coins:) and receive(coins:)—to handle
custom class to open a file and write some data to it, you might need to close the file the distribution and collection of coins.
before the class instance is deallocated.
The distribute(coins:) method checks that there are enough coins in the bank
Class definitions can have at most one deinitializer per class. The deinitializer does before distributing them. If there are not enough coins, Bank returns a smaller number
not take any parameters and is written without parentheses: than the number that was requested (and returns zero if no coins are left in the bank).
It returns an integer value to indicate the actual number of coins that were provided.
deinit {
// perform the deinitialization The receive(coins:) method simply adds the received number of coins back into the
} bank’s coin store.
Deinitializers are called automatically, just before instance deallocation takes place.
The Player class describes a player in the game. Each player has a certain number
You are not allowed to call a deinitializer yourself. Superclass deinitializers are
of coins stored in their purse at any time. This is represented by the player’s
inherited by their subclasses, and the superclass deinitializer is called automatically at
coinsInPurse property:
the end of a subclass deinitializer implementation. Superclass deinitializers are
always called, even if a subclass does not provide its own deinitializer.
class Player {
var coinsInPurse: Int
Because an instance is not deallocated until after its deinitializer is called, a
deinitializer can access all properties of the instance it is called on and can modify its init(coins: Int) {
behavior based on those properties (such as looking up the name of a file that needs coinsInPurse = Bank.distribute(coins: coins)
to be closed). }
func win(coins: Int) {
Deinitializers in Action coinsInPurse += Bank.distribute(coins: coins)
Here’s an example of a deinitializer in action. This example defines two new types, }
Bank and Player, for a simple game. The Bank class manages a made-up currency,
deinit {

118
 Bank.receive(coins: coinsInPurse) The player has now left the game. This is indicated by setting the optional playerOne
} variable to nil, meaning “no Player instance.” At the point that this happens, the
} playerOne variable’s reference to the Player instance is broken. No other properties
or variables are still referring to the Player instance, and so it is deallocated in order
Each Player instance is initialized with a starting allowance of a specified number of to free up its memory. Just before this happens, its deinitializer is called automatically,
coins from the bank during initialization, although a Player instance may receive fewer and its coins are returned to the bank.
than that number if not enough coins are available.

The Player class defines a win(coins:) method, which retrieves a certain number of
coins from the bank and adds them to the player’s purse. The Player class also
implements a deinitializer, which is called just before a Player instance is deallocated.
Here, the deinitializer simply returns all of the player’s coins to the bank:

var playerOne: Player? = Player(coins: 100)

print("A new player has joined the game with \
(playerOne!.coinsInPurse) coins")
// Prints "A new player has joined the game with 100 coins"
print("There are now \(Bank.coinsInBank) coins left in the bank")
// Prints "There are now 9900 coins left in the bank"

A new Player instance is created, with a request for 100 coins if they are available.
This Player instance is stored in an optional Player variable called playerOne. An
optional variable is used here, because players can leave the game at any point. The
optional lets you track whether there is currently a player in the game.

Because playerOne is an optional, it is qualified with an exclamation mark (!) when its
coinsInPurse property is accessed to print its default number of coins, and whenever
its winCoins(_:) method is called:

playerOne!.win(coins: 2_000)
print("PlayerOne won 2000 coins & now has \
(playerOne!.coinsInPurse) coins")
// Prints "PlayerOne won 2000 coins & now has 2100 coins"
print("The bank now only has \(Bank.coinsInBank) coins left")
// Prints "The bank now only has 7900 coins left"

Here, the player has won 2,000 coins. The player’s purse now contains 2,100 coins,
and the bank has only 7,900 coins left.

playerOne = nil
print("PlayerOne has left the game")
// Prints "PlayerOne has left the game"
print("The bank now has \(Bank.coinsInBank) coins")
// Prints "The bank now has 10000 coins"

119
 Section 16 To make this possible, whenever you assign a class instance to a property, constant,
or variable, that property, constant, or variable makes a strong reference to the

Automatic Reference instance. The reference is called a “strong” reference because it keeps a firm hold on
that instance, and does not allow it to be deallocated for as long as that strong
reference remains.

ARC in Action
Automatic Reference Counting Here’s an example of how Automatic Reference Counting works. This example starts
with a simple class called Person, which defines a stored constant property called
Swift uses Automatic Reference Counting (ARC) to track and manage your app’s name:
memory usage. In most cases, this means that memory management “just works” in
Swift, and you do not need to think about memory management yourself. ARC class Person {
automatically frees up the memory used by class instances when those instances are let name: String
no longer needed.
init(name: String) {
self.name = name
However, in a few cases ARC requires more information about the relationships
between parts of your code in order to manage memory for you. This chapter print("\(name) is being initialized")
describes those situations and shows how you enable ARC to manage all of your }
app’s memory. Using ARC in Swift is very similar to the approach described in deinit {
Transitioning to ARC Release Notes for using ARC with Objective-C.
print("\(name) is being deinitialized")

NOTE }
}
Reference counting only applies to instances of classes. Structures and enumerations
are value types, not reference types, and are not stored and passed by reference. The Person class has an initializer that sets the instance’s name property and prints a
message to indicate that initialization is underway. The Person class also has a
How ARC Works deinitializer that prints a message when an instance of the class is deallocated.
Every time you create a new instance of a class, ARC allocates a chunk of memory to
store information about that instance. This memory holds information about the type of The next code snippet defines three variables of type Person?, which are used to set
the instance, together with the values of any stored properties associated with that up multiple references to a new Person instance in subsequent code snippets.
instance. Because these variables are of an optional type (Person?, not Person), they are
automatically initialized with a value of nil, and do not currently reference a Person
Additionally, when an instance is no longer needed, ARC frees up the memory used instance.
by that instance so that the memory can be used for other purposes instead. This
var reference1: Person?
ensures that class instances do not take up space in memory when they are no longer
needed. var reference2: Person?
var reference3: Person?
However, if ARC were to deallocate an instance that was still in use, it would no You can now create a new Person instance and assign it to one of these three
longer be possible to access that instance’s properties, or call that instance’s variables:
methods. Indeed, if you tried to access the instance, your app would most likely crash.
reference1 = Person(name: "John Appleseed")
To make sure that instances don’t disappear while they are still needed, ARC tracks // Prints "John Appleseed is being initialized"
how many properties, constants, and variables are currently referring to each class
instance. ARC will not deallocate an instance as long as at least one active reference Note that the message "John Appleseed is being initialized" is printed at the
to that instance still exists. point that you call the Person class’s initializer. This confirms that initialization has
taken place.

120
Because the new Person instance has been assigned to the reference1 variable, class Person {
there is now a strong reference from reference1 to the new Person instance. Because let name: String
there is at least one strong reference, ARC makes sure that this Person is kept in init(name: String) { self.name = name }
memory and is not deallocated.
var apartment: Apartment?
deinit { print("\(name) is being deinitialized") }
If you assign the same Person instance to two more variables, two more strong
references to that instance are established: }

reference2 = reference1
class Apartment {
reference3 = reference1
let unit: String
There are now three strong references to this single Person instance. init(unit: String) { self.unit = unit }
var tenant: Person?
If you break two of these strong references (including the original reference) by
deinit { print("Apartment \(unit) is being deinitialized") }
assigning nil to two of the variables, a single strong reference remains, and the
Person instance is not deallocated: }

Every Person instance has a name property of type String and an optional apartment
reference1 = nil property that is initially nil. The apartment property is optional, because a person
reference2 = nil may not always have an apartment.
ARC does not deallocate the Person instance until the third and final strong reference
is broken, at which point it is clear that you are no longer using the Person instance: Similarly, every Apartment instance has a unit property of type String and has an
optional tenant property that is initially nil. The tenant property is optional because
reference3 = nil an apartment may not always have a tenant.
// Prints "John Appleseed is being deinitialized"
Both of these classes also define a deinitializer, which prints the fact that an instance
of that class is being deinitialized. This enables you to see whether instances of
Strong Reference Cycles Between Class Instances Person and Apartment are being deallocated as expected.
In the examples above, ARC is able to track the number of references to the new
Person instance you create and to deallocate that Person instance when it is no longer This next code snippet defines two variables of optional type called john and unit4A,
needed. which will be set to a specific Apartment and Person instance below. Both of these
variables have an initial value of nil, by virtue of being optional:
However, it is possible to write code in which an instance of a class never gets to a
point where it has zero strong references. This can happen if two class instances hold var john: Person?
a strong reference to each other, such that each instance keeps the other alive. This is var unit4A: Apartment?
known as a strong reference cycle.
You can now create a specific Person instance and Apartment instance and assign
You resolve strong reference cycles by defining some of the relationships between these new instances to the john and unit4A variables:
classes as weak or unowned references instead of as strong references. This process
john = Person(name: "John Appleseed")
is described in Resolving Strong Reference Cycles Between Class Instances.
However, before you learn how to resolve a strong reference cycle, it is useful to unit4A = Apartment(unit: "4A")
understand how such a cycle is caused. Here’s how the strong references look after creating and assigning these two
instances. The john variable now has a strong reference to the new Person instance,
Here’s an example of how a strong reference cycle can be created by accident. This and the unit4A variable has a strong reference to the new Apartment instance:
example defines two classes called Person and Apartment, which model a block of
apartments and its residents:

121
You can now link the two instances together so that the person has an apartment, and The strong references between the Person instance and the Apartment instance
the apartment has a tenant. Note that an exclamation mark (!) is used to unwrap and remain and cannot be broken.
access the instances stored inside the john and unit4A optional variables, so that the
properties of those instances can be set:
Resolving Strong Reference Cycles Between Class Instances
john!.apartment = unit4A Swift provides two ways to resolve strong reference cycles when you work with
unit4A!.tenant = john properties of class type: weak references and unowned references.

Here’s how the strong references look after you link the two instances together:
Weak and unowned references enable one instance in a reference cycle to refer to
the other instance without keeping a strong hold on it. The instances can then refer to
each other without creating a strong reference cycle.

Use a weak reference when the other instance has a shorter lifetime—that is, when
the other instance can be deallocated first. In the Apartment example above, it is
appropriate for an apartment to be able to have no tenant at some point in its lifetime,
and so a weak reference is an appropriate way to break the reference cycle in this
case. In contrast, use an unowned reference when the other instance has the same
lifetime or a longer lifetime.
Unfortunately, linking these two instances creates a strong reference cycle between
them. The Person instance now has a strong reference to the Apartment instance, and Weak References
the Apartment instance has a strong reference to the Person instance. Therefore, A weak reference is a reference that does not keep a strong hold on the instance it
when you break the strong references held by the john and unit4A variables, the refers to, and so does not stop ARC from disposing of the referenced instance. This
reference counts do not drop to zero, and the instances are not deallocated by ARC: behavior prevents the reference from becoming part of a strong reference cycle. You
indicate a weak reference by placing the weak keyword before a property or variable
john = nil declaration.
unit4A = nil

Note that neither deinitializer was called when you set these two variables to nil. The Because a weak reference does not keep a strong hold on the instance it refers to, it
strong reference cycle prevents the Person and Apartment instances from ever being is possible for that instance to be deallocated while the weak reference is still referring
deallocated, causing a memory leak in your app. to it. Therefore, ARC automatically sets a weak reference to nil when the instance
that it refers to is deallocated. And, because weak references need to allow their value
Here’s how the strong references look after you set the john and unit4A variables to to be changed to nil at runtime, they are always declared as variables, rather than
nil: constants, of an optional type.

You can check for the existence of a value in the weak reference, just like any other
optional value, and you will never end up with a reference to an invalid instance that
no longer exists.

NOTE

122
 Property observers aren’t called when ARC sets a weak reference to nil. The Person instance still has a strong reference to the Apartment instance, but the
Apartment instance now has a weak reference to the Person instance. This means
The example below is identical to the Person and Apartment example from above,
that when you break the strong reference held by the john variable by setting it to nil,
with one important difference. This time around, the Apartment type’s tenant property
there are no more strong references to the Person instance:
is declared as a weak reference:
john = nil
class Person {
// Prints "John Appleseed is being deinitialized"
let name: String
init(name: String) { self.name = name } Because there are no more strong references to the Person instance, it is deallocated
and the tenant property is set to nil:
var apartment: Apartment?
deinit { print("\(name) is being deinitialized") }
}

class Apartment {
let unit: String
init(unit: String) { self.unit = unit }
weak var tenant: Person?
deinit { print("Apartment \(unit) is being deinitialized") } The only remaining strong reference to the Apartment instance is from the unit4A
} variable. If you break that strong reference, there are no more strong references to the
Apartment instance:
The strong references from the two variables (john and unit4A) and the links between
the two instances are created as before: unit4A = nil

var john: Person? // Prints "Apartment 4A is being deinitialized"

var unit4A: Apartment? Because there are no more strong references to the Apartment instance, it too is
deallocated:
john = Person(name: "John Appleseed")
unit4A = Apartment(unit: "4A")

john!.apartment = unit4A
unit4A!.tenant = john

Here’s how the references look now that you’ve linked the two instances together:

NOTE

In systems that use garbage collection, weak pointers are sometimes used to
implement a simple caching mechanism because objects with no strong references are
deallocated only when memory pressure triggers garbage collection. However, with
ARC, values are deallocated as soon as their last strong reference is removed, making
weak references unsuitable for such a purpose.

Unowned References

123
Like a weak reference, an unowned reference does not keep a strong hold on the }
instance it refers to. Unlike a weak reference, however, an unowned reference is used
when the other instance has the same lifetime or a longer lifetime. You indicate an class CreditCard {
unowned reference by placing the unowned keyword before a property or variable
let number: UInt64
declaration.
unowned let customer: Customer

An unowned reference is expected to always have a value. As a result, ARC never init(number: UInt64, customer: Customer) {
sets an unowned reference’s value to nil, which means that unowned references are self.number = number
defined using nonoptional types. self.customer = customer
}
I M P O R TA N T
deinit { print("Card #\(number) is being deinitialized") }
Use an unowned reference only when you are sure that the reference always refers to
an instance that has not been deallocated. }
NOTE
If you try to access the value of an unowned reference after that instance has been
deallocated, you’ll get a runtime error. The number property of the CreditCard class is defined with a type of UInt64 rather
than Int, to ensure that the number property’s capacity is large enough to store a 16-
The following example defines two classes, Customer and CreditCard, which model a digit card number on both 32-bit and 64-bit systems.
bank customer and a possible credit card for that customer. These two classes each
store an instance of the other class as a property. This relationship has the potential to This next code snippet defines an optional Customer variable called john, which will
create a strong reference cycle. be used to store a reference to a specific customer. This variable has an initial value
of nil, by virtue of being optional:
The relationship between Customer and CreditCard is slightly different from the
var john: Customer?
relationship between Apartment and Person seen in the weak reference example
above. In this data model, a customer may or may not have a credit card, but a credit You can now create a Customer instance, and use it to initialize and assign a new
card will always be associated with a customer. A CreditCard instance never outlives CreditCard instance as that customer’s card property:
the Customer that it refers to. To represent this, the Customer class has an optional
card property, but the CreditCard class has an unowned (and nonoptional) customer john = Customer(name: "John Appleseed")
property. john!.card = CreditCard(number: 1234_5678_9012_3456, customer:
john!)
Furthermore, a new CreditCard instance can only be created by passing a number Here’s how the references look, now that you’ve linked the two instances:
value and a customer instance to a custom CreditCard initializer. This ensures that a
CreditCard instance always has a customer instance associated with it when the
CreditCard instance is created.

Because a credit card will always have a customer, you define its customer property
as an unowned reference, to avoid a strong reference cycle:

class Customer {
let name: String
var card: CreditCard? The Customer instance now has a strong reference to the CreditCard instance, and
init(name: String) { the CreditCard instance has an unowned reference to the Customer instance.
self.name = name
} Because of the unowned customer reference, when you break the strong reference
held by the john variable, there are no more strong references to the Customer
deinit { print("\(name) is being deinitialized") }
instance:

124
 This enables both properties to be accessed directly (without optional unwrapping)
once initialization is complete, while still avoiding a reference cycle. This section
shows you how to set up such a relationship.

The example below defines two classes, Country and City, each of which stores an
instance of the other class as a property. In this data model, every country must
always have a capital city, and every city must always belong to a country. To
represent this, the Country class has a capitalCity property, and the City class has
Because there are no more strong references to the Customer instance, it is a country property:
deallocated. After this happens, there are no more strong references to the
CreditCard instance, and it too is deallocated: class Country {
let name: String
john = nil
var capitalCity: City!
// Prints "John Appleseed is being deinitialized"
init(name: String, capitalName: String) {
// Prints "Card #1234567890123456 is being deinitialized"
self.name = name
The final code snippet above shows that the deinitializers for the Customer instance
self.capitalCity = City(name: capitalName, country: self)
and CreditCard instance both print their “deinitialized” messages after the john
variable is set to nil. }
}
NOTE

The examples above show how to use safe unowned references. Swift also provides class City {
unsafe unowned references for cases where you need to disable runtime safety checks
—for example, for performance reasons. As with all unsafe operations, you take on the let name: String
responsiblity for checking that code for safety. unowned let country: Country

You indicate an unsafe unowned reference by writing unowned(unsafe). If you try to init(name: String, country: Country) {
access an unsafe unowned reference after the instance that it refers to is deallocated, self.name = name
your program will try to access the memory location where the instance used to be,
which is an unsafe operation. self.country = country
}
Unowned References and Implicitly Unwrapped Optional Properties }
The examples for weak and unowned references above cover two of the more
To set up the interdependency between the two classes, the initializer for City takes a
common scenarios in which it is necessary to break a strong reference cycle.
Country instance, and stores this instance in its country property.

The Person and Apartment example shows a situation where two properties, both of
The initializer for City is called from within the initializer for Country. However, the
which are allowed to be nil, have the potential to cause a strong reference cycle. This
initializer for Country cannot pass self to the City initializer until a new Country
scenario is best resolved with a weak reference.
instance is fully initialized, as described in Two-Phase Initialization.

The Customer and CreditCard example shows a situation where one property that is
To cope with this requirement, you declare the capitalCity property of Country as an
allowed to be nil and another property that cannot be nil have the potential to cause
implicitly unwrapped optional property, indicated by the exclamation mark at the end of
a strong reference cycle. This scenario is best resolved with an unowned reference.
its type annotation (City!). This means that the capitalCity property has a default
value of nil, like any other optional, but can be accessed without the need to unwrap
However, there is a third scenario, in which both properties should always have a its value as described in Implicitly Unwrapped Optionals.
value, and neither property should ever be nil once initialization is complete. In this
scenario, it is useful to combine an unowned property on one class with an implicitly
Because capitalCity has a default nil value, a new Country instance is considered
unwrapped optional property on the other class.
fully initialized as soon as the Country instance sets its name property within its

125
initializer. This means that the Country initializer can start to reference and pass class HTMLElement {
around the implicit self property as soon as the name property is set. The Country
initializer can therefore pass self as one of the parameters for the City initializer let name: String
when the Country initializer is setting its own capitalCity property.
let text: String?

All of this means that you can create the Country and City instances in a single
statement, without creating a strong reference cycle, and the capitalCity property lazy var asHTML: () -> String = {
can be accessed directly, without needing to use an exclamation mark to unwrap its if let text = self.text {
optional value: return "<\(self.name)>\(text)</\(self.name)>"
} else {
var country = Country(name: "Canada", capitalName: "Ottawa")
return "<\(self.name) />"
print("\(country.name)'s capital city is called \
(country.capitalCity.name)") }
// Prints "Canada's capital city is called Ottawa" }

In the example above, the use of an implicitly unwrapped optional means that all of
the two-phase class initializer requirements are satisfied. The capitalCity property init(name: String, text: String? = nil) {
can be used and accessed like a nonoptional value once initialization is complete, self.name = name
while still avoiding a strong reference cycle.
self.text = text
}
Strong Reference Cycles for Closures
You saw above how a strong reference cycle can be created when two class instance
deinit {
properties hold a strong reference to each other. You also saw how to use weak and
unowned references to break these strong reference cycles. print("\(name) is being deinitialized")
}
A strong reference cycle can also occur if you assign a closure to a property of a class
instance, and the body of that closure captures the instance. This capture might occur }
because the closure’s body accesses a property of the instance, such as
self.someProperty, or because the closure calls a method on the instance, such as The HTMLElement class defines a name property, which indicates the name of the
self.someMethod(). In either case, these accesses cause the closure to “capture” element, such as "h1" for a heading element, "p" for a paragraph element, or "br" for
self, creating a strong reference cycle. a line break element. HTMLElement also defines an optional text property, which you
can set to a string that represents the text to be rendered within that HTML element.
This strong reference cycle occurs because closures, like classes, are reference
types. When you assign a closure to a property, you are assigning a reference to that In addition to these two simple properties, the HTMLElement class defines a lazy
closure. In essence, it’s the same problem as above—two strong references are property called asHTML. This property references a closure that combines name and
keeping each other alive. However, rather than two class instances, this time it’s a text into an HTML string fragment. The asHTML property is of type () -> String, or “a
class instance and a closure that are keeping each other alive. function that takes no parameters, and returns a String value”.

Swift provides an elegant solution to this problem, known as a closure capture list. By default, the asHTML property is assigned a closure that returns a string
However, before you learn how to break a strong reference cycle with a closure representation of an HTML tag. This tag contains the optional text value if it exists, or
capture list, it is useful to understand how such a cycle can be caused. no text content if text does not exist. For a paragraph element, the closure would
return "<p>some text</p>" or "<p />", depending on whether the text property
The example below shows how you can create a strong reference cycle when using a equals "some text" or nil.
closure that references self. This example defines a class called HTMLElement, which
provides a simple model for an individual element within an HTML document:

126
The asHTML property is named and used somewhat like an instance method. However,
because asHTML is a closure property rather than an instance method, you can replace
the default value of the asHTML property with a custom closure, if you want to change
the HTML rendering for a particular HTML element.

For example, the asHTML property could be set to a closure that defaults to some text if
the text property is nil, in order to prevent the representation from returning an
empty HTML tag:
The instance’s asHTML property holds a strong reference to its closure. However,
let heading = HTMLElement(name: "h1")
because the closure refers to self within its body (as a way to reference self.name
let defaultText = "some default text" and self.text), the closure captures self, which means that it holds a strong
heading.asHTML = { reference back to the HTMLElement instance. A strong reference cycle is created
return "<\(heading.name)>\(heading.text ?? defaultText)</\ between the two. (For more information about capturing values in a closure, see
(heading.name)>" Capturing Values.)
}
NOTE
print(heading.asHTML())
Even though the closure refers to self multiple times, it only captures one strong
// Prints "<h1>some default text</h1>"
reference to the HTMLElement instance.
NOTE
If you set the paragraph variable to nil and break its strong reference to the
The asHTML property is declared as a lazy property, because it is only needed if and
HTMLElement instance, neither the HTMLElement instance nor its closure are
when the element actually needs to be rendered as a string value for some HTML
output target. The fact that asHTML is a lazy property means that you can refer to self deallocated, because of the strong reference cycle:
within the default closure, because the lazy property will not be accessed until after
initialization has been completed and self is known to exist. paragraph = nil

The HTMLElement class provides a single initializer, which takes a name argument and Note that the message in the HTMLElement deinitializer is not printed, which shows that
(if desired) a text argument to initialize a new element. The class also defines a the HTMLElement instance is not deallocated.
deinitializer, which prints a message to show when an HTMLElement instance is
deallocated. Resolving Strong Reference Cycles for Closures
You resolve a strong reference cycle between a closure and a class instance by
Here’s how you use the HTMLElement class to create and print a new instance: defining a capture list as part of the closure’s definition. A capture list defines the rules
to use when capturing one or more reference types within the closure’s body. As with
var paragraph: HTMLElement? = HTMLElement(name: "p", text: "hello, strong reference cycles between two class instances, you declare each captured
world")
reference to be a weak or unowned reference rather than a strong reference. The
print(paragraph!.asHTML()) appropriate choice of weak or unowned depends on the relationships between the
// Prints "<p>hello, world</p>" different parts of your code.
NOTE
NOTE
The paragraph variable above is defined as an optional HTMLElement, so that it can be
set to nil below to demonstrate the presence of a strong reference cycle. Swift requires you to write self.someProperty or self.someMethod() (rather than
just someProperty or someMethod()) whenever you refer to a member of self within a
Unfortunately, the HTMLElement class, as written above, creates a strong reference closure. This helps you remember that it’s possible to capture self by accident.
cycle between an HTMLElement instance and the closure used for its default asHTML
value. Here’s how the cycle looks: Defining a Capture List
Each item in a capture list is a pairing of the weak or unowned keyword with a reference
to a class instance (such as self) or a variable initialized with some value (such as

127
delegate = self.delegate!). These pairings are written within a pair of square lazy var asHTML: () -> String = {
braces, separated by commas. [unowned self] in
if let text = self.text {
Place the capture list before a closure’s parameter list and return type if they are return "<\(self.name)>\(text)</\(self.name)>"
provided:
} else {
lazy var someClosure: (Int, String) -> String = { return "<\(self.name) />"
[unowned self, weak delegate = self.delegate!] (index: Int, }
stringToProcess: String) -> String in
}
// closure body goes here
}
init(name: String, text: String? = nil) {
If a closure does not specify a parameter list or return type because they can be self.name = name
inferred from context, place the capture list at the very start of the closure, followed by self.text = text
the in keyword:
}
lazy var someClosure: () -> String = {
[unowned self, weak delegate = self.delegate!] in deinit {
// closure body goes here print("\(name) is being deinitialized")
} }

Weak and Unowned References

}
Define a capture in a closure as an unowned reference when the closure and the
instance it captures will always refer to each other, and will always be deallocated at This implementation of HTMLElement is identical to the previous implementation, apart
the same time. from the addition of a capture list within the asHTML closure. In this case, the capture
list is [unowned self], which means “capture self as an unowned reference rather
Conversely, define a capture as a weak reference when the captured reference may than a strong reference”.
become nil at some point in the future. Weak references are always of an optional
type, and automatically become nil when the instance they reference is deallocated. You can create and print an HTMLElement instance as before:
This enables you to check for their existence within the closure’s body.
var paragraph: HTMLElement? = HTMLElement(name: "p", text: "hello,
world")
NOTE
print(paragraph!.asHTML())
If the captured reference will never become nil, it should always be captured as an
// Prints "<p>hello, world</p>"
unowned reference, rather than a weak reference.
Here’s how the references look with the capture list in place:
An unowned reference is the appropriate capture method to use to resolve the strong
reference cycle in the HTMLElement example from earlier. Here’s how you write the
HTMLElement class to avoid the cycle:

class HTMLElement {

let name: String

let text: String?

128
This time, the capture of self by the closure is an unowned reference, and does not
keep a strong hold on the HTMLElement instance it has captured. If you set the strong
reference from the paragraph variable to nil, the HTMLElement instance is
deallocated, as can be seen from the printing of its deinitializer message in the
example below:

paragraph = nil
// Prints "p is being deinitialized"

For more information about capture lists, see Capture Lists.

129
 var residence: Residence?
Section 17 }

Optional Chaining class Residence {

var numberOfRooms = 1
}

Optional Chaining Residence instances have a single Int property called numberOfRooms, with a default
value of 1. Person instances have an optional residence property of type Residence?.
Optional chaining is a process for querying and calling properties, methods, and
subscripts on an optional that might currently be nil. If the optional contains a value, If you create a new Person instance, its residence property is default initialized to nil,
the property, method, or subscript call succeeds; if the optional is nil, the property, by virtue of being optional. In the code below, john has a residence property value of
method, or subscript call returns nil. Multiple queries can be chained together, and nil:
the entire chain fails gracefully if any link in the chain is nil.
let john = Person()
NOTE
If you try to access the numberOfRooms property of this person’s residence, by placing
Optional chaining in Swift is similar to messaging nil in Objective-C, but in a way that an exclamation mark after residence to force the unwrapping of its value, you trigger
works for any type, and that can be checked for success or failure. a runtime error, because there is no residence value to unwrap:

Optional Chaining as an Alternative to Forced Unwrapping let roomCount = john.residence!.numberOfRooms

You specify optional chaining by placing a question mark (?) after the optional value // this triggers a runtime error
on which you wish to call a property, method or subscript if the optional is non-nil. The code above succeeds when john.residence has a non-nil value and will set
This is very similar to placing an exclamation mark (!) after an optional value to force roomCount to an Int value containing the appropriate number of rooms. However, this
the unwrapping of its value. The main difference is that optional chaining fails code always triggers a runtime error when residence is nil, as illustrated above.
gracefully when the optional is nil, whereas forced unwrapping triggers a runtime
error when the optional is nil. Optional chaining provides an alternative way to access the value of numberOfRooms.
To use optional chaining, use a question mark in place of the exclamation mark:
To reflect the fact that optional chaining can be called on a nil value, the result of an
optional chaining call is always an optional value, even if the property, method, or if let roomCount = john.residence?.numberOfRooms {
subscript you are querying returns a nonoptional value. You can use this optional print("John's residence has \(roomCount) room(s).")
return value to check whether the optional chaining call was successful (the returned
} else {
optional contains a value), or did not succeed due to a nil value in the chain (the
returned optional value is nil). print("Unable to retrieve the number of rooms.")
}
Specifically, the result of an optional chaining call is of the same type as the expected // Prints "Unable to retrieve the number of rooms."
return value, but wrapped in an optional. A property that normally returns an Int will
This tells Swift to “chain” on the optional residence property and to retrieve the value
return an Int? when accessed through optional chaining.
of numberOfRooms if residence exists.

The next several code snippets demonstrate how optional chaining differs from forced
Because the attempt to access numberOfRooms has the potential to fail, the optional
unwrapping and enables you to check for success.
chaining attempt returns a value of type Int?, or “optional Int”. When residence is
nil, as in the example above, this optional Int will also be nil, to reflect the fact that
First, two classes called Person and Residence are defined: it was not possible to access numberOfRooms. The optional Int is accessed through
optional binding to unwrap the integer and assign the nonoptional value to the
class Person {
roomCount variable.

130
Note that this is true even though numberOfRooms is a nonoptional Int. The fact that it return rooms.count
is queried through an optional chain means that the call to numberOfRooms will always }
return an Int? instead of an Int. subscript(i: Int) -> Room {
get {
You can assign a Residence instance to john.residence, so that it no longer has a
return rooms[i]
nil value:
}
john.residence = Residence() set {
john.residence now contains an actual Residence instance, rather than nil. If you try rooms[i] = newValue
to access numberOfRooms with the same optional chaining as before, it will now return }
an Int? that contains the default numberOfRooms value of 1: }

if let roomCount = john.residence?.numberOfRooms { func printNumberOfRooms() {

print("John's residence has \(roomCount) room(s).") print("The number of rooms is \(numberOfRooms)")

} else { }

print("Unable to retrieve the number of rooms.") var address: Address?

} }

// Prints "John's residence has 1 room(s)." Because this version of Residence stores an array of Room instances, its
numberOfRooms property is implemented as a computed property, not a stored
Defining Model Classes for Optional Chaining property. The computed numberOfRooms property simply returns the value of the count
property from the rooms array.
You can use optional chaining with calls to properties, methods, and subscripts that
are more than one level deep. This enables you to drill down into subproperties within
As a shortcut to accessing its rooms array, this version of Residence provides a read-
complex models of interrelated types, and to check whether it is possible to access
write subscript that provides access to the room at the requested index in the rooms
properties, methods, and subscripts on those subproperties.
array.
The code snippets below define four model classes for use in several subsequent
This version of Residence also provides a method called printNumberOfRooms, which
examples, including examples of multilevel optional chaining. These classes expand
simply prints the number of rooms in the residence.
upon the Person and Residence model from above by adding a Room and Address
class, with associated properties, methods, and subscripts.
Finally, Residence defines an optional property called address, with a type of
Address?. The Address class type for this property is defined below.
The Person class is defined in the same way as before:

class Person { The Room class used for the rooms array is a simple class with one property called
name, and an initializer to set that property to a suitable room name:
var residence: Residence?
} class Room {
The Residence class is more complex than before. This time, the Residence class let name: String
defines a variable property called rooms, which is initialized with an empty array of init(name: String) { self.name = name }
type [Room]:
}

class Residence { The final class in this model is called Address. This class has three optional properties
var rooms = [Room]() of type String?. The first two properties, buildingName and buildingNumber, are
alternative ways to identify a particular building as part of an address. The third
var numberOfRooms: Int {
property, street, is used to name the street for that address:

131
 class Address { someAddress.buildingNumber = "29"
var buildingName: String? someAddress.street = "Acacia Road"
var buildingNumber: String? john.residence?.address = someAddress
var street: String? In this example, the attempt to set the address property of john.residence will fail,
func buildingIdentifier() -> String? { because john.residence is currently nil.
if buildingNumber != nil && street != nil {
return "\(buildingNumber) \(street)"
The assignment is part of the optional chaining, which means none of the code on the
right hand side of the = operator is evaluated. In the previous example, it’s not easy to
} else if buildingName != nil {
see that someAddress is never evaluated, because accessing a constant doesn’t have
return buildingName any side effects. The listing below does the same assignment, but it uses a function to
} else { create the address. The function prints “Function was called” before returning a value,
return nil which lets you see whether the right hand side of the = operator was evaluated.
}
func createAddress() -> Address {
}
print("Function was called.")
}

The Address class also provides a method called buildingIdentifier(), which has a let someAddress = Address()
return type of String?. This method checks the properties of the address and returns
someAddress.buildingNumber = "29"
buildingName if it has a value, or buildingNumber concatenated with street if both
have values, or nil otherwise. someAddress.street = "Acacia Road"

Accessing Properties Through Optional Chaining return someAddress

As demonstrated in Optional Chaining as an Alternative to Forced Unwrapping, you }

can use optional chaining to access a property on an optional value, and to check if john.residence?.address = createAddress()
that property access is successful.
You can tell that the createAddress() function isn’t called, because nothing is printed.

Use the classes defined above to create a new Person instance, and try to access its
numberOfRooms property as before:
Calling Methods Through Optional Chaining
You can use optional chaining to call a method on an optional value, and to check
let john = Person() whether that method call is successful. You can do this even if that method does not
if let roomCount = john.residence?.numberOfRooms { define a return value.
print("John's residence has \(roomCount) room(s).")
The printNumberOfRooms() method on the Residence class prints the current value of
} else {
numberOfRooms. Here’s how the method looks:
print("Unable to retrieve the number of rooms.")
} func printNumberOfRooms() {
// Prints "Unable to retrieve the number of rooms." print("The number of rooms is \(numberOfRooms)")

Because john.residence is nil, this optional chaining call fails in the same way as }
before. This method does not specify a return type. However, functions and methods with no
return type have an implicit return type of Void, as described in Functions Without
You can also attempt to set a property’s value through optional chaining: Return Values. This means that they return a value of (), or an empty tuple.

let someAddress = Address()

132
If you call this method on an optional value with optional chaining, the method’s return print("Unable to retrieve the first room name.")
type will be Void?, not Void, because return values are always of an optional type }
when called through optional chaining. This enables you to use an if statement to // Prints "Unable to retrieve the first room name."
check whether it was possible to call the printNumberOfRooms() method, even though
the method does not itself define a return value. Compare the return value from the The optional chaining question mark in this subscript call is placed immediately after
john.residence, before the subscript brackets, because john.residence is the
printNumberOfRooms call against nil to see if the method call was successful:
optional value on which optional chaining is being attempted.
if john.residence?.printNumberOfRooms() != nil {
print("It was possible to print the number of rooms.")
Similarly, you can try to set a new value through a subscript with optional chaining:
} else { john.residence?[0] = Room(name: "Bathroom")
print("It was not possible to print the number of rooms.")
This subscript setting attempt also fails, because residence is currently nil.
}
// Prints "It was not possible to print the number of rooms." If you create and assign an actual Residence instance to john.residence, with one or
The same is true if you attempt to set a property through optional chaining. The more Room instances in its rooms array, you can use the Residence subscript to access
example above in Accessing Properties Through Optional Chaining attempts to set an the actual items in the rooms array through optional chaining:
address value for john.residence, even though the residence property is nil. Any
let johnsHouse = Residence()
attempt to set a property through optional chaining returns a value of type Void?,
which enables you to compare against nil to see if the property was set successfully: johnsHouse.rooms.append(Room(name: "Living Room"))
johnsHouse.rooms.append(Room(name: "Kitchen"))
if (john.residence?.address = someAddress) != nil {
john.residence = johnsHouse
print("It was possible to set the address.")
} else {
if let firstRoomName = john.residence?[0].name {
print("It was not possible to set the address.")
print("The first room name is \(firstRoomName).")
}
} else {
// Prints "It was not possible to set the address."
print("Unable to retrieve the first room name.")
}
Accessing Subscripts Through Optional Chaining
// Prints "The first room name is Living Room."
You can use optional chaining to try to retrieve and set a value from a subscript on an
optional value, and to check whether that subscript call is successful. Accessing Subscripts of Optional Type
If a subscript returns a value of optional type—such as the key subscript of Swift’s
NOTE
Dictionary type—place a question mark after the subscript’s closing bracket to chain
When you access a subscript on an optional value through optional chaining, you place on its optional return value:
the question mark before the subscript’s brackets, not after. The optional chaining
question mark always follows immediately after the part of the expression that is var testScores = ["Dave": [86, 82, 84], "Bev": [79, 94, 81]]
optional.
testScores["Dave"]?[0] = 91
The example below tries to retrieve the name of the first room in the rooms array of the testScores["Bev"]?[0] += 1
john.residence property using the subscript defined on the Residence class.
testScores["Brian"]?[0] = 72
Because john.residence is currently nil, the subscript call fails:
// the "Dave" array is now [91, 82, 84] and the "Bev" array is now
[80, 94, 81]
if let firstRoomName = john.residence?[0].name {
print("The first room name is \(firstRoomName).") The example above defines a dictionary called testScores, which contains two key-
} else {
value pairs that map a String key to an array of Int values. The example uses

133
optional chaining to set the first item in the "Dave" array to 91; to increment the first of optional chaining are applied in addition to the underlying optional type of the
item in the "Bev" array by 1; and to try to set the first item in an array for a key of property.
"Brian". The first two calls succeed, because the testScores dictionary contains keys
for "Dave" and "Bev". The third call fails, because the testScores dictionary does not If you set an actual Address instance as the value for john.residence.address, and
contain a key for "Brian". set an actual value for the address’s street property, you can access the value of the
street property through multilevel optional chaining:
Linking Multiple Levels of Chaining
let johnsAddress = Address()
You can link together multiple levels of optional chaining to drill down to properties,
methods, and subscripts deeper within a model. However, multiple levels of optional johnsAddress.buildingName = "The Larches"
chaining do not add more levels of optionality to the returned value. johnsAddress.street = "Laurel Street"
john.residence?.address = johnsAddress
To put it another way:
if let johnsStreet = john.residence?.address?.street {
If the type you are trying to retrieve is not optional, it will become optional
print("John's street name is \(johnsStreet).")
because of the optional chaining.
} else {
If the type you are trying to retrieve is already optional, it will not become
print("Unable to retrieve the address.")
more optional because of the chaining.
}
Therefore:
// Prints "John's street name is Laurel Street."

If you try to retrieve an Int value through optional chaining, an Int? is In this example, the attempt to set the address property of john.residence will
always returned, no matter how many levels of chaining are used. succeed, because the value of john.residence currently contains a valid Residence
instance.
Similarly, if you try to retrieve an Int? value through optional chaining, an
Int? is always returned, no matter how many levels of chaining are used.
Chaining on Methods with Optional Return Values
The example below tries to access the street property of the address property of the
The previous example shows how to retrieve the value of a property of optional type
residence property of john. There are two levels of optional chaining in use here, to
through optional chaining. You can also use optional chaining to call a method that
chain through the residence and address properties, both of which are of optional
returns a value of optional type, and to chain on that method’s return value if needed.
type:

if let johnsStreet = john.residence?.address?.street { The example below calls the Address class’s buildingIdentifier() method through
optional chaining. This method returns a value of type String?. As described above,
print("John's street name is \(johnsStreet).")
the ultimate return type of this method call after optional chaining is also String?:
} else {
print("Unable to retrieve the address.") if let buildingIdentifier =
john.residence?.address?.buildingIdentifier() {
}
print("John's building identifier is \(buildingIdentifier).")
// Prints "Unable to retrieve the address."
}
The value of john.residence currently contains a valid Residence instance. However,
// Prints "John's building identifier is The Larches."
the value of john.residence.address is currently nil. Because of this, the call to
john.residence?.address?.street fails. If you want to perform further optional chaining on this method’s return value, place
the optional chaining question mark after the method’s parentheses:
Note that in the example above, you are trying to retrieve the value of the street
property. The type of this property is String?. The return value of if let beginsWithThe =
john.residence?.address?.street is therefore also String?, even though two levels john.residence?.address?.buildingIdentifier()?.hasPrefix("The")
{

134
 if beginsWithThe {
print("John's building identifier begins with \"The\".")
} else {
print("John's building identifier does not begin with
\"The\".")
}
}
// Prints "John's building identifier begins with "The"."
NOTE

In the example above, you place the optional chaining question mark after the
parentheses, because the optional value you are chaining on is the
buildingIdentifier() method’s return value, and not the buildingIdentifier()
method itself.

135
 Section 18 Throwing an error lets you indicate that something unexpected happened and the
normal flow of execution can’t continue. You use a throw statement to throw an error.

Error Handling For example, the following code throws an error to indicate that five additional coins
are needed by the vending machine:

throw VendingMachineError.insufficientFunds(coinsNeeded: 5)

Error Handling Handling Errors

When an error is thrown, some surrounding piece of code must be responsible for
Error handling is the process of responding to and recovering from error conditions in handling the error—for example, by correcting the problem, trying an alternative
your program. Swift provides first-class support for throwing, catching, propagating, approach, or informing the user of the failure.
and manipulating recoverable errors at runtime.
There are four ways to handle errors in Swift. You can propagate the error from a
Some operations aren’t guaranteed to always complete execution or produce a useful function to the code that calls that function, handle the error using a do-catch
output. Optionals are used to represent the absence of a value, but when an statement, handle the error as an optional value, or assert that the error will not occur.
operation fails, it’s often useful to understand what caused the failure, so that your Each approach is described in a section below.
code can respond accordingly.
When a function throws an error, it changes the flow of your program, so it’s important
As an example, consider the task of reading and processing data from a file on disk. that you can quickly identify places in your code that can throw errors. To identify
There are a number of ways this task can fail, including the file not existing at the these places in your code, write the try keyword—or the try? or try! variation—
specified path, the file not having read permissions, or the file not being encoded in a before a piece of code that calls a function, method, or initializer that can throw an
compatible format. Distinguishing among these different situations allows a program error. These keywords are described in the sections below.
to resolve some errors and to communicate to the user any errors it can’t resolve.
NOTE
NOTE
Error handling in Swift resembles exception handling in other languages, with the use
Error handling in Swift interoperates with error handling patterns that use the NSError of the try, catch and throw keywords. Unlike exception handling in many languages
class in Cocoa and Objective-C. For more information about this class, see Error —including Objective-C—error handling in Swift does not involve unwinding the call
Handling in Using Swift with Cocoa and Objective-C (Swift 3.0.1). stack, a process that can be computationally expensive. As such, the performance
characteristics of a throw statement are comparable to those of a return statement.

Representing and Throwing Errors Propagating Errors Using Throwing Functions

In Swift, errors are represented by values of types that conform to the Error protocol. To indicate that a function, method, or initializer can throw an error, you write the
This empty protocol indicates that a type can be used for error handling. throws keyword in the function’s declaration after its parameters. A function marked
with throws is called a throwing function. If the function specifies a return type, you
Swift enumerations are particularly well suited to modeling a group of related error write the throws keyword before the return arrow (->).
conditions, with associated values allowing for additional information about the nature
of an error to be communicated. For example, here’s how you might represent the func canThrowErrors() throws -> String
error conditions of operating a vending machine inside a game:

enum VendingMachineError: Error { func cannotThrowErrors() -> String

case invalidSelection A throwing function propagates errors that are thrown inside of it to the scope from
case insufficientFunds(coinsNeeded: Int)
which it’s called.
case outOfStock NOTE
}

136
 Only throwing functions can propagate errors. Any errors thrown inside a nonthrowing print("Dispensing \(name)")
function must be handled inside the function. }
In the example below, the VendingMachine class has a vend(itemNamed:) method that }
throws an appropriate VendingMachineError if the requested item is not available, is The implementation of the vend(itemNamed:) method uses guard statements to exit
out of stock, or has a cost that exceeds the current deposited amount: the method early and throw appropriate errors if any of the requirements for
purchasing a snack aren’t met. Because a throw statement immediately transfers
struct Item {
program control, an item will be vended only if all of these requirements are met.
var price: Int
var count: Int Because the vend(itemNamed:) method propagates any errors it throws, any code
} that calls this method must either handle the errors—using a do-catch statement,
try?, or try!—or continue to propagate them. For example, the
buyFavoriteSnack(person:vendingMachine:) in the example below is also a throwing
class VendingMachine {
function, and any errors that the vend(itemNamed:) method throws will propagate up
var inventory = [ to the point where the buyFavoriteSnack(person:vendingMachine:) function is
"Candy Bar": Item(price: 12, count: 7), called.
"Chips": Item(price: 10, count: 4),
let favoriteSnacks = [
"Pretzels": Item(price: 7, count: 11)
"Alice": "Chips",
]
"Bob": "Licorice",
var coinsDeposited = 0
"Eve": "Pretzels",
]
func vend(itemNamed name: String) throws {
func buyFavoriteSnack(person: String, vendingMachine:
guard let item = inventory[name] else { VendingMachine) throws {
throw VendingMachineError.invalidSelection let snackName = favoriteSnacks[person] ?? "Candy Bar"
} try vendingMachine.vend(itemNamed: snackName)
}
guard item.count > 0 else {
In this example, the buyFavoriteSnack(person: vendingMachine:) function looks up
throw VendingMachineError.outOfStock a given person’s favorite snack and tries to buy it for them by calling the
} vend(itemNamed:) method. Because the vend(itemNamed:) method can throw an
error, it’s called with the try keyword in front of it.
guard item.price <= coinsDeposited else {
Throwing initializers can propagate errors in the same way as throwing functions. For
throw VendingMachineError.insufficientFunds(coinsNeeded:
item.price - coinsDeposited) example, the initializer for the PurchasedSnack structure in the listing below calls a
throwing function as part of the initialization process, and it handles any errors that it
}
encounters by propagating them to its caller.

coinsDeposited -= item.price struct PurchasedSnack {

let name: String
var newItem = item init(name: String, vendingMachine: VendingMachine) throws {
newItem.count -= 1 try vendingMachine.vend(itemNamed: name)
inventory[name] = newItem self.name = name
}

137
 } In the above example, the buyFavoriteSnack(person:vendingMachine:) function is
called in a try expression, because it can throw an error. If an error is thrown,
Handling Errors Using Do-Catch execution immediately transfers to the catch clauses, which decide whether to allow
You use a do-catch statement to handle errors by running a block of code. If an error propagation to continue. If no error is thrown, the remaining statements in the do
is thrown by the code in the do clause, it is matched against the catch clauses to statement are executed.
determine which one of them can handle the error.
Converting Errors to Optional Values
Here is the general form of a do-catch statement: You use try? to handle an error by converting it to an optional value. If an error is
thrown while evaluating the try? expression, the value of the expression is nil. For
do {
try expression example, in the following code x and y have the same value and behavior:
statements
} catch pattern 1 { func someThrowingFunction() throws -> Int {
statements
// ...
} catch pattern 2 where condition {
statements }
}
You write a pattern after catch to indicate what errors that clause can handle. If a
let x = try? someThrowingFunction()
catch clause doesn’t have a pattern, the clause matches any error and binds the error
to a local constant named error. For more information about pattern matching, see
Patterns. let y: Int?
do {
The catch clauses don’t have to handle every possible error that the code in its do y = try someThrowingFunction()
clause can throw. If none of the catch clauses handle the error, the error propagates
} catch {
to the surrounding scope. However, the error must be handled by some surrounding
scope—either by an enclosing do-catch clause that handles the error or by being y = nil
inside a throwing function. For example, the following code handles all three cases of }
the VendingMachineError enumeration, but all other errors have to be handled by its If someThrowingFunction() throws an error, the value of x and y is nil. Otherwise, the
surrounding scope: value of x and y is the value that the function returned. Note that x and y are an
optional of whatever type someThrowingFunction() returns. Here the function returns
var vendingMachine = VendingMachine()
an integer, so x and y are optional integers.
vendingMachine.coinsDeposited = 8
do { Using try? lets you write concise error handling code when you want to handle all
try buyFavoriteSnack(person: "Alice", vendingMachine: errors in the same way. For example, the following code uses several approaches to
vendingMachine) fetch data, or returns nil if all of the approaches fail.
} catch VendingMachineError.invalidSelection {
func fetchData() -> Data? {
print("Invalid Selection.")
if let data = try? fetchDataFromDisk() { return data }
} catch VendingMachineError.outOfStock {
if let data = try? fetchDataFromServer() { return data }
print("Out of Stock.")
return nil
} catch VendingMachineError.insufficientFunds(let coinsNeeded) {
}
print("Insufficient funds. Please insert an additional \
(coinsNeeded) coins.")
Disabling Error Propagation
}
Sometimes you know a throwing function or method won’t, in fact, throw an error at
// Prints "Insufficient funds. Please insert an additional 2 coins."
runtime. On those occasions, you can write try! before the expression to disable

138
error propagation and wrap the call in a runtime assertion that no error will be thrown. You can use a defer statement even when no error handling code is involved.
If an error actually is thrown, you’ll get a runtime error.

For example, the following code uses a loadImage(atPath:) function, which loads the
image resource at a given path or throws an error if the image can’t be loaded. In this
case, because the image is shipped with the application, no error will be thrown at
runtime, so it is appropriate to disable error propagation.

let photo = try! loadImage(atPath: "./Resources/John

Appleseed.jpg")

Specifying Cleanup Actions

You use a defer statement to execute a set of statements just before code execution
leaves the current block of code. This statement lets you do any necessary cleanup
that should be performed regardless of how execution leaves the current block of
code—whether it leaves because an error was thrown or because of a statement such
as return or break. For example, you can use a defer statement to ensure that file
descriptors are closed and manually allocated memory is freed.

A defer statement defers execution until the current scope is exited. This statement
consists of the defer keyword and the statements to be executed later. The deferred
statements may not contain any code that would transfer control out of the
statements, such as a break or a return statement, or by throwing an error. Deferred
actions are executed in reverse order of how they are specified—that is, the code in
the first defer statement executes after code in the second, and so on.

func processFile(filename: String) throws {

if exists(filename) {
let file = open(filename)
defer {
close(file)
}
while let line = try file.readline() {
// Work with the file.
}
// close(file) is called here, at the end of the scope.
}
}

The above example uses a defer statement to ensure that the open(_:) function has
a corresponding call to close(_:).

NOTE

139
 init(name: String, director: String) {
Section 19 self.director = director

Type Casting }
super.init(name: name)

Type Casting class Song: MediaItem {

var artist: String
Type casting is a way to check the type of an instance, or to treat that instance as a init(name: String, artist: String) {
different superclass or subclass from somewhere else in its own class hierarchy. self.artist = artist
super.init(name: name)
Type casting in Swift is implemented with the is and as operators. These two
operators provide a simple and expressive way to check the type of a value or cast a }
value to a different type. }

The final snippet creates a constant array called library, which contains two Movie
You can also use type casting to check whether a type conforms to a protocol, as instances and three Song instances. The type of the library array is inferred by
described in Checking for Protocol Conformance. initializing it with the contents of an array literal. Swift’s type checker is able to deduce
that Movie and Song have a common superclass of MediaItem, and so it infers a type
Defining a Class Hierarchy for Type Casting of [MediaItem] for the library array:
You can use type casting with a hierarchy of classes and subclasses to check the type
of a particular class instance and to cast that instance to another class within the let library = [
same hierarchy. The three code snippets below define a hierarchy of classes and an Movie(name: "Casablanca", director: "Michael Curtiz"),
array containing instances of those classes, for use in an example of type casting. Song(name: "Blue Suede Shoes", artist: "Elvis Presley"),
Movie(name: "Citizen Kane", director: "Orson Welles"),
The first snippet defines a new base class called MediaItem. This class provides basic
Song(name: "The One And Only", artist: "Chesney Hawkes"),
functionality for any kind of item that appears in a digital media library. Specifically, it
declares a name property of type String, and an init name initializer. (It is assumed Song(name: "Never Gonna Give You Up", artist: "Rick Astley")
that all media items, including all movies and songs, will have a name.) ]
// the type of "library" is inferred to be [MediaItem]
class MediaItem {
The items stored in library are still Movie and Song instances behind the scenes.
var name: String
However, if you iterate over the contents of this array, the items you receive back are
init(name: String) { typed as MediaItem, and not as Movie or Song. In order to work with them as their
self.name = name native type, you need to check their type, or downcast them to a different type, as
} described below.
}
Checking Type
The next snippet defines two subclasses of MediaItem. The first subclass, Movie,
encapsulates additional information about a movie or film. It adds a director property Use the type check operator (is) to check whether an instance is of a certain subclass
on top of the base MediaItem class, with a corresponding initializer. The second type. The type check operator returns true if the instance is of that subclass type and
false if it is not.
subclass, Song, adds an artist property and initializer on top of the base class:

class Movie: MediaItem { The example below defines two variables, movieCount and songCount, which count
var director: String
the number of Movie and Song instances in the library array:

140
 var movieCount = 0 The example below iterates over each MediaItem in library, and prints an
var songCount = 0 appropriate description for each item. To do this, it needs to access each item as a
true Movie or Song, and not just as a MediaItem. This is necessary in order for it to be
able to access the director or artist property of a Movie or Song for use in the
for item in library {
description.
if item is Movie {
movieCount += 1 In this example, each item in the array might be a Movie, or it might be a Song. You
} else if item is Song { don’t know in advance which actual class to use for each item, and so it is appropriate
songCount += 1 to use the conditional form of the type cast operator (as?) to check the downcast each
time through the loop:
}
} for item in library {
if let movie = item as? Movie {
print("Media library contains \(movieCount) movies and \(songCount) print("Movie: \(movie.name), dir. \(movie.director)")
songs")
} else if let song = item as? Song {
// Prints "Media library contains 2 movies and 3 songs"
print("Song: \(song.name), by \(song.artist)")
This example iterates through all items in the library array. On each pass, the for-in
}
loop sets the item constant to the next MediaItem in the array.
}

item is Movie returns true if the current MediaItem is a Movie instance and false if it
is not. Similarly, item is Song checks whether the item is a Song instance. At the end // Movie: Casablanca, dir. Michael Curtiz
of the for-in loop, the values of movieCount and songCount contain a count of how // Song: Blue Suede Shoes, by Elvis Presley
many MediaItem instances were found of each type.
// Movie: Citizen Kane, dir. Orson Welles
// Song: The One And Only, by Chesney Hawkes
Downcasting
// Song: Never Gonna Give You Up, by Rick Astley
A constant or variable of a certain class type may actually refer to an instance of a
subclass behind the scenes. Where you believe this is the case, you can try to The example starts by trying to downcast the current item as a Movie. Because item
downcast to the subclass type with a type cast operator (as? or as!). is a MediaItem instance, it’s possible that it might be a Movie; equally, it’s also possible
that it might be a Song, or even just a base MediaItem. Because of this uncertainty, the
Because downcasting can fail, the type cast operator comes in two different forms. as? form of the type cast operator returns an optional value when attempting to
The conditional form, as?, returns an optional value of the type you are trying to downcast to a subclass type. The result of item as? Movie is of type Movie?, or
downcast to. The forced form, as!, attempts the downcast and force-unwraps the “optional Movie”.
result as a single compound action.
Downcasting to Movie fails when applied to the Song instances in the library array. To
Use the conditional form of the type cast operator (as?) when you are not sure if the cope with this, the example above uses optional binding to check whether the optional
downcast will succeed. This form of the operator will always return an optional value, Movie actually contains a value (that is, to find out whether the downcast succeeded.)
and the value will be nil if the downcast was not possible. This enables you to check This optional binding is written “if let movie = item as? Movie”, which can be read
for a successful downcast. as:

Use the forced form of the type cast operator (as!) only when you are sure that the “Try to access item as a Movie. If this is successful, set a new temporary constant
downcast will always succeed. This form of the operator will trigger a runtime error if called movie to the value stored in the returned optional Movie.”
you try to downcast to an incorrect class type.
If the downcasting succeeds, the properties of movie are then used to print a
description for that Movie instance, including the name of its director. A similar

141
principle is used to check for Song instances, and to print an appropriate description for thing in things {
(including artist name) whenever a Song is found in the library. switch thing {
case 0 as Int:
NOTE
print("zero as an Int")
Casting does not actually modify the instance or change its values. The underlying
case 0 as Double:
instance remains the same; it is simply treated and accessed as an instance of the type
to which it has been cast. print("zero as a Double")
case let someInt as Int:
Type Casting for Any and AnyObject print("an integer value of \(someInt)")
Swift provides two special types for working with nonspecific types: case let someDouble as Double where someDouble > 0:
print("a positive double value of \(someDouble)")
Any can represent an instance of any type at all, including function types.
case is Double:
AnyObject can represent an instance of any class type. print("some other double value that I don't want to print")
Use Any and AnyObject only when you explicitly need the behavior and capabilities case let someString as String:
they provide. It is always better to be specific about the types you expect to work with print("a string value of \"\(someString)\"")
in your code.
case let (x, y) as (Double, Double):
print("an (x, y) point at \(x), \(y)")
Here’s an example of using Any to work with a mix of different types, including function
types and non-class types. The example creates an array called things, which can case let movie as Movie:
store values of type Any: print("a movie called \(movie.name), dir. \
(movie.director)")
var things = [Any]() case let stringConverter as (String) -> String:
print(stringConverter("Michael"))
things.append(0) default:
things.append(0.0) print("something else")
things.append(42) }
things.append(3.14159) }
things.append("hello")
things.append((3.0, 5.0)) // zero as an Int
things.append(Movie(name: "Ghostbusters", director: "Ivan // zero as a Double
Reitman"))
// an integer value of 42
things.append({ (name: String) -> String in "Hello, \(name)" })
// a positive double value of 3.14159
The things array contains two Int values, two Double values, a String value, a tuple // a string value of "hello"
of type (Double, Double), the movie “Ghostbusters”, and a closure expression that
// an (x, y) point at 3.0, 5.0
takes a String value and returns another String value.
// a movie called Ghostbusters, dir. Ivan Reitman
To discover the specific type of a constant or variable that is known only to be of type // Hello, Michael
Any or AnyObject, you can use an is or as pattern in a switch statement’s cases. The NOTE
example below iterates over the items in the things array and queries the type of
The Any type represents values of any type, including optional types. Swift gives you a
each item with a switch statement. Several of the switch statement’s cases bind their
warning if you use an optional value where a value of type Any is expected. If you really
matched value to a constant of the specified type to enable its value to be printed: do need to use an optional value as an Any value, you can use the as operator to
explicitly cast the optional to Any, as shown below.

142
let optionalNumber: Int? = 3
things.append(optionalNumber) // Warning
things.append(optionalNumber as Any) // No warning

143
 }
Section 20 var values: Values {

Nested Types switch self {

case .ace:
return Values(first: 1, second: 11)
case .jack, .queen, .king:

Nested Types return Values(first: 10, second: nil)

default:
Enumerations are often created to support a specific class or structure’s functionality. return Values(first: self.rawValue, second: nil)
Similarly, it can be convenient to define utility classes and structures purely for use }
within the context of a more complex type. To accomplish this, Swift enables you to
}
define nested types, whereby you nest supporting enumerations, classes, and
structures within the definition of the type they support. }

To nest a type within another type, write its definition within the outer braces of the // BlackjackCard properties and methods
type it supports. Types can be nested to as many levels as are required. let rank: Rank, suit: Suit
var description: String {
Nested Types in Action
var output = "suit is \(suit.rawValue),"
The example below defines a structure called BlackjackCard, which models a playing
output += " value is \(rank.values.first)"
card as used in the game of Blackjack. The BlackJack structure contains two nested
enumeration types called Suit and Rank. if let second = rank.values.second {
output += " or \(second)"
In Blackjack, the Ace cards have a value of either one or eleven. This feature is }
represented by a structure called Values, which is nested within the Rank return output
enumeration:
}
struct BlackjackCard { }

The Suit enumeration describes the four common playing card suits, together with a
// nested Suit enumeration raw Character value to represent their symbol.
enum Suit: Character {
The Rank enumeration describes the thirteen possible playing card ranks, together
case spades = "♠", hearts = "♡", diamonds = "♢", clubs =
"♣" with a raw Int value to represent their face value. (This raw Int value is not used for
the Jack, Queen, King, and Ace cards.)
}

As mentioned above, the Rank enumeration defines a further nested structure of its
// nested Rank enumeration own, called Values. This structure encapsulates the fact that most cards have one
enum Rank: Int { value, but the Ace card has two values. The Values structure defines two properties to
case two = 2, three, four, five, six, seven, eight, nine, represent this:
ten
case jack, queen, king, ace first, of type Int

struct Values { second, of type Int?, or “optional Int”

let first: Int, second: Int?

144
Rank also defines a computed property, values, which returns an instance of the
Values structure. This computed property considers the rank of the card and initializes
a new Values instance with appropriate values based on its rank. It uses special
values for jack, queen, king, and ace. For the numeric cards, it uses the rank’s raw
Int value.

The BlackjackCard structure itself has two properties—rank and suit. It also defines
a computed property called description, which uses the values stored in rank and
suit to build a description of the name and value of the card. The description
property uses optional binding to check whether there is a second value to display,
and if so, inserts additional description detail for that second value.

Because BlackjackCard is a structure with no custom initializers, it has an implicit

memberwise initializer, as described in Memberwise Initializers for Structure Types.
You can use this initializer to initialize a new constant called theAceOfSpades:

let theAceOfSpades = BlackjackCard(rank: .ace, suit: .spades)

print("theAceOfSpades: \(theAceOfSpades.description)")
// Prints "theAceOfSpades: suit is ♠, value is 1 or 11"

Even though Rank and Suit are nested within BlackjackCard, their type can be
inferred from context, and so the initialization of this instance is able to refer to the
enumeration cases by their case names (.ace and .spades) alone. In the example
above, the description property correctly reports that the Ace of Spades has a value
of 1 or 11.

Referring to Nested Types

To use a nested type outside of its definition context, prefix its name with the name of
the type it is nested within:

let heartsSymbol = BlackjackCard.Suit.hearts.rawValue

// heartsSymbol is "♡"

For the example above, this enables the names of Suit, Rank, and Values to be kept
deliberately short, because their names are naturally qualified by the context in which
they are defined.

145
 extension SomeType: SomeProtocol, AnotherProtocol {
Section 21 // implementation of protocol requirements goes here

Extensions }

Adding protocol conformance in this way is described in Adding Protocol

Conformance with an Extension.

NOTE
Extensions
If you define an extension to add new functionality to an existing type, the new
functionality will be available on all existing instances of that type, even if they were
Extensions add new functionality to an existing class, structure, enumeration, or created before the extension was defined.
protocol type. This includes the ability to extend types for which you do not have
access to the original source code (known as retroactive modeling). Extensions are
similar to categories in Objective-C. (Unlike Objective-C categories, Swift extensions
Computed Properties
do not have names.) Extensions can add computed instance properties and computed type properties to
existing types. This example adds five computed instance properties to Swift’s built-in
Extensions in Swift can: Double type, to provide basic support for working with distance units:

extension Double {
Add computed instance properties and computed type properties
var km: Double { return self * 1_000.0 }
Define instance methods and type methods
var m: Double { return self }
Provide new initializers var cm: Double { return self / 100.0 }
Define subscripts var mm: Double { return self / 1_000.0 }

Define and use new nested types var ft: Double { return self / 3.28084 }
}
Make an existing type conform to a protocol
let oneInch = 25.4.mm
In Swift, you can even extend a protocol to provide implementations of its
print("One inch is \(oneInch) meters")
requirements or add additional functionality that conforming types can take advantage
of. For more details, see Protocol Extensions. // Prints "One inch is 0.0254 meters"
let threeFeet = 3.ft
NOTE
print("Three feet is \(threeFeet) meters")
Extensions can add new functionality to a type, but they cannot override existing // Prints "Three feet is 0.914399970739201 meters"
functionality.
These computed properties express that a Double value should be considered as a
certain unit of length. Although they are implemented as computed properties, the
Extension Syntax
names of these properties can be appended to a floating-point literal value with dot
Declare extensions with the extension keyword: syntax, as a way to use that literal value to perform distance conversions.
extension SomeType {
In this example, a Double value of 1.0 is considered to represent “one meter”. This is
// new functionality to add to SomeType goes here why the m computed property returns self—the expression 1.m is considered to
} calculate a Double value of 1.0.
An extension can extend an existing type to make it adopt one or more protocols.
Where this is the case, the protocol names are written in exactly the same way as for Other units require some conversion to be expressed as a value measured in meters.
a class or structure: One kilometer is the same as 1,000 meters, so the km computed property multiplies
the value by 1_000.00 to convert into a number expressed in meters. Similarly, there

146
are 3.28084 feet in a meter, and so the ft computed property divides the underlying var origin = Point()
Double value by 3.28084, to convert it from feet to meters. var size = Size()
}
These properties are read-only computed properties, and so they are expressed
Because the Rect structure provides default values for all of its properties, it receives
without the get keyword, for brevity. Their return value is of type Double, and can be
a default initializer and a memberwise initializer automatically, as described in Default
used within mathematical calculations wherever a Double is accepted:
Initializers. These initializers can be used to create new Rect instances:
let aMarathon = 42.km + 195.m
let defaultRect = Rect()
print("A marathon is \(aMarathon) meters long")
let memberwiseRect = Rect(origin: Point(x: 2.0, y: 2.0),
// Prints "A marathon is 42195.0 meters long"
size: Size(width: 5.0, height: 5.0))
NOTE
You can extend the Rect structure to provide an additional initializer that takes a
Extensions can add new computed properties, but they cannot add stored properties, specific center point and size:
or add property observers to existing properties.
extension Rect {
Initializers init(center: Point, size: Size) {
Extensions can add new initializers to existing types. This enables you to extend other let originX = center.x - (size.width / 2)
types to accept your own custom types as initializer parameters, or to provide
let originY = center.y - (size.height / 2)
additional initialization options that were not included as part of the type’s original
implementation. self.init(origin: Point(x: originX, y: originY), size:
size)

Extensions can add new convenience initializers to a class, but they cannot add new }
designated initializers or deinitializers to a class. Designated initializers and }
deinitializers must always be provided by the original class implementation. This new initializer starts by calculating an appropriate origin point based on the
provided center point and size value. The initializer then calls the structure’s
NOTE
automatic memberwise initializer init(origin:size:), which stores the new origin
If you use an extension to add an initializer to a value type that provides default values and size values in the appropriate properties:
for all of its stored properties and does not define any custom initializers, you can call
the default initializer and memberwise initializer for that value type from within your let centerRect = Rect(center: Point(x: 4.0, y: 4.0),
extension’s initializer.
size: Size(width: 3.0, height: 3.0))
This would not be the case if you had written the initializer as part of the value type’s // centerRect's origin is (2.5, 2.5) and its size is (3.0, 3.0)
original implementation, as described in Initializer Delegation for Value Types.
NOTE
The example below defines a custom Rect structure to represent a geometric
If you provide a new initializer with an extension, you are still responsible for making
rectangle. The example also defines two supporting structures called Size and Point,
sure that each instance is fully initialized once the initializer completes.
both of which provide default values of 0.0 for all of their properties:

struct Size { Methods

var width = 0.0, height = 0.0 Extensions can add new instance methods and type methods to existing types. The
following example adds a new instance method called repetitions to the Int type:
}
struct Point { extension Int {
var x = 0.0, y = 0.0 func repetitions(task: () -> Void) {
} for _ in 0..<self {
struct Rect { task()

147
 } …and so on:
}
extension Int {
}
subscript(digitIndex: Int) -> Int {
The repetitions(task:) method takes a single argument of type () -> Void, which
indicates a function that has no parameters and does not return a value. var decimalBase = 1
for _ in 0..<digitIndex {
After defining this extension, you can call the repetitions(task:) method on any decimalBase *= 10
integer to perform a task that many number of times: }
return (self / decimalBase) % 10
3.repetitions {
}
print("Hello!")
}
}
746381295[0]
// Hello!
// returns 5
// Hello!
746381295[1]
// Hello!
// returns 9
Mutating Instance Methods 746381295[2]
Instance methods added with an extension can also modify (or mutate) the instance // returns 2
itself. Structure and enumeration methods that modify self or its properties must mark 746381295[8]
the instance method as mutating, just like mutating methods from an original
// returns 7
implementation.
If the Int value does not have enough digits for the requested index, the subscript
The example below adds a new mutating method called square to Swift’s Int type, implementation returns 0, as if the number had been padded with zeros to the left:
which squares the original value:
746381295[9]
extension Int { // returns 0, as if you had requested:
mutating func square() { 0746381295[9]
self = self * self
} Nested Types
} Extensions can add new nested types to existing classes, structures, and
var someInt = 3 enumerations:
someInt.square()
extension Int {
// someInt is now 9
enum Kind {
case negative, zero, positive
Subscripts
}
Extensions can add new subscripts to an existing type. This example adds an integer
var kind: Kind {
subscript to Swift’s built-in Int type. This subscript [n] returns the decimal digit n
places in from the right of the number: switch self {
case 0:
123456789[0] returns 9 return .zero
123456789[1] returns 8 case let x where x > 0:

148
 return .positive
default:
return .negative
}
}
}

This example adds a new nested enumeration to Int. This enumeration, called Kind,
expresses the kind of number that a particular integer represents. Specifically, it
expresses whether the number is negative, zero, or positive.

This example also adds a new computed instance property to Int, called kind, which
returns the appropriate Kind enumeration case for that integer.

The nested enumeration can now be used with any Int value:

func printIntegerKinds(_ numbers: [Int]) {

for number in numbers {
switch number.kind {
case .negative:
print("- ", terminator: "")
case .zero:
print("0 ", terminator: "")
case .positive:
print("+ ", terminator: "")
}
}
print("")
}
printIntegerKinds([3, 19, -27, 0, -6, 0, 7])
// Prints "+ + - 0 - 0 + "

This function, printIntegerKinds(_:), takes an input array of Int values and iterates
over those values in turn. For each integer in the array, the function considers the
kind computed property for that integer, and prints an appropriate description.

NOTE

number.kind is already known to be of type Int.Kind. Because of this, all of the

Int.Kind case values can be written in shorthand form inside the switch statement,
such as .negative rather than Int.Kind.negative.

149
 Section 22 property should be a stored property or a computed property—it only specifies the
required property name and type. The protocol also specifies whether each property

Protocols must be gettable or gettable and settable.

If a protocol requires a property to be gettable and settable, that property requirement

cannot be fulfilled by a constant stored property or a read-only computed property. If
the protocol only requires a property to be gettable, the requirement can be satisfied
Protocols by any kind of property, and it is valid for the property to be also settable if this is
useful for your own code.
A protocol defines a blueprint of methods, properties, and other requirements that suit
a particular task or piece of functionality. The protocol can then be adopted by a class, Property requirements are always declared as variable properties, prefixed with the
structure, or enumeration to provide an actual implementation of those requirements. var keyword. Gettable and settable properties are indicated by writing { get set }
Any type that satisfies the requirements of a protocol is said to conform to that after their type declaration, and gettable properties are indicated by writing { get }.
protocol.
protocol SomeProtocol {

In addition to specifying requirements that conforming types must implement, you can var mustBeSettable: Int { get set }
extend a protocol to implement some of these requirements or to implement additional var doesNotNeedToBeSettable: Int { get }
functionality that conforming types can take advantage of. }

Always prefix type property requirements with the static keyword when you define
Protocol Syntax them in a protocol. This rule pertains even though type property requirements can be
You define protocols in a very similar way to classes, structures, and enumerations: prefixed with the class or static keyword when implemented by a class:

protocol SomeProtocol { protocol AnotherProtocol {

// protocol definition goes here static var someTypeProperty: Int { get set }
} }
Custom types state that they adopt a particular protocol by placing the protocol’s Here’s an example of a protocol with a single instance property requirement:
name after the type’s name, separated by a colon, as part of their definition. Multiple
protocols can be listed, and are separated by commas: protocol FullyNamed {
var fullName: String { get }
struct SomeStructure: FirstProtocol, AnotherProtocol {
}
// structure definition goes here
The FullyNamed protocol requires a conforming type to provide a fully-qualified name.
}
The protocol doesn’t specify anything else about the nature of the conforming type—it
If a class has a superclass, list the superclass name before any protocols it adopts, only specifies that the type must be able to provide a full name for itself. The protocol
followed by a comma: states that any FullyNamed type must have a gettable instance property called
fullName, which is of type String.
class SomeClass: SomeSuperclass, FirstProtocol, AnotherProtocol {
// class definition goes here Here’s an example of a simple structure that adopts and conforms to the FullyNamed
} protocol:

struct Person: FullyNamed {

Property Requirements
var fullName: String
A protocol can require any conforming type to provide an instance property or type
property with a particular name and type. The protocol doesn’t specify whether the }
let john = Person(fullName: "John Appleseed")

150
 // john.fullName is "John Appleseed" method requirements are prefixed with the class or static keyword when
This example defines a structure called Person, which represents a specific named implemented by a class:
person. It states that it adopts the FullyNamed protocol as part of the first line of its
definition. protocol SomeProtocol {
static func someTypeMethod()
Each instance of Person has a single stored property called fullName, which is of type }
String. This matches the single requirement of the FullyNamed protocol, and means
The following example defines a protocol with a single instance method requirement:
that Person has correctly conformed to the protocol. (Swift reports an error at compile-
time if a protocol requirement is not fulfilled.) protocol RandomNumberGenerator {
func random() -> Double
Here’s a more complex class, which also adopts and conforms to the FullyNamed
protocol: }

This protocol, RandomNumberGenerator, requires any conforming type to have an

class Starship: FullyNamed { instance method called random, which returns a Double value whenever it is called.
var prefix: String? Although it is not specified as part of the protocol, it is assumed that this value will be
var name: String a number from 0.0 up to (but not including) 1.0.
init(name: String, prefix: String? = nil) {
The RandomNumberGenerator protocol does not make any assumptions about how
self.name = name
each random number will be generated—it simply requires the generator to provide a
self.prefix = prefix standard way to generate a new random number.
}
var fullName: String { Here’s an implementation of a class that adopts and conforms to the
RandomNumberGenerator protocol. This class implements a pseudorandom number
return (prefix != nil ? prefix! + " " : "") + name
generator algorithm known as a linear congruential generator:
}
} class LinearCongruentialGenerator: RandomNumberGenerator {
var ncc1701 = Starship(name: "Enterprise", prefix: "USS") var lastRandom = 42.0
// ncc1701.fullName is "USS Enterprise" let m = 139968.0
This class implements the fullName property requirement as a computed read-only let a = 3877.0
property for a starship. Each Starship class instance stores a mandatory name and an let c = 29573.0
optional prefix. The fullName property uses the prefix value if it exists, and func random() -> Double {
prepends it to the beginning of name to create a full name for the starship.
lastRandom = ((lastRandom * a +
c).truncatingRemainder(dividingBy:m))
Method Requirements return lastRandom / m
Protocols can require specific instance methods and type methods to be implemented }
by conforming types. These methods are written as part of the protocol’s definition in
}
exactly the same way as for normal instance and type methods, but without curly
braces or a method body. Variadic parameters are allowed, subject to the same rules let generator = LinearCongruentialGenerator()
as for normal methods. Default values, however, cannot be specified for method print("Here's a random number: \(generator.random())")
parameters within a protocol’s definition. // Prints "Here's a random number: 0.37464991998171"
print("And another one: \(generator.random())")
As with type property requirements, you always prefix type method requirements with
// Prints "And another one: 0.729023776863283"
the static keyword when they are defined in a protocol. This is true even though type

151
 self = .on
Mutating Method Requirements
case .on:
It is sometimes necessary for a method to modify (or mutate) the instance it belongs
to. For instance methods on value types (that is, structures and enumerations) you self = .off
place the mutating keyword before a method’s func keyword to indicate that the }
method is allowed to modify the instance it belongs to and any properties of that }
instance. This process is described in Modifying Value Types from Within Instance
}
Methods.
var lightSwitch = OnOffSwitch.off

If you define a protocol instance method requirement that is intended to mutate lightSwitch.toggle()
instances of any type that adopts the protocol, mark the method with the mutating // lightSwitch is now equal to .on
keyword as part of the protocol’s definition. This enables structures and enumerations
to adopt the protocol and satisfy that method requirement. Initializer Requirements
Protocols can require specific initializers to be implemented by conforming types. You
NOTE
write these initializers as part of the protocol’s definition in exactly the same way as for
If you mark a protocol instance method requirement as mutating, you do not need to normal initializers, but without curly braces or an initializer body:
write the mutating keyword when writing an implementation of that method for a class.
The mutating keyword is only used by structures and enumerations. protocol SomeProtocol {
The example below defines a protocol called Togglable, which defines a single init(someParameter: Int)
instance method requirement called toggle. As its name suggests, the toggle() }
method is intended to toggle or invert the state of any conforming type, typically by
modifying a property of that type. Class Implementations of Protocol Initializer Requirements
You can implement a protocol initializer requirement on a conforming class as either a
The toggle() method is marked with the mutating keyword as part of the Togglable designated initializer or a convenience initializer. In both cases, you must mark the
protocol definition, to indicate that the method is expected to mutate the state of a initializer implementation with the required modifier:
conforming instance when it is called:
class SomeClass: SomeProtocol {
protocol Togglable {
required init(someParameter: Int) {
mutating func toggle()
// initializer implementation goes here
}
}
If you implement the Togglable protocol for a structure or enumeration, that structure }
or enumeration can conform to the protocol by providing an implementation of the
toggle() method that is also marked as mutating.
The use of the required modifier ensures that you provide an explicit or inherited
implementation of the initializer requirement on all subclasses of the conforming class,
such that they also conform to the protocol.
The example below defines an enumeration called OnOffSwitch. This enumeration
toggles between two states, indicated by the enumeration cases on and off. The
enumeration’s toggle implementation is marked as mutating, to match the Togglable For more information on required initializers, see Required Initializers.
protocol’s requirements:
NOTE

enum OnOffSwitch: Togglable { You do not need to mark protocol initializer implementations with the required modifier
case off, on on classes that are marked with the final modifier, because final classes cannot be
subclassed. For more on the final modifier, see Preventing Overrides.
mutating func toggle() {
switch self {
case .off:

152
If a subclass overrides a designated initializer from a superclass, and also implements Because protocols are types, begin their names with a capital letter (such as
a matching initializer requirement from a protocol, mark the initializer implementation FullyNamed and RandomNumberGenerator) to match the names of other types in Swift
(such as Int, String, and Double).
with both the required and override modifiers:
Here’s an example of a protocol used as a type:
protocol SomeProtocol {
init() class Dice {
} let sides: Int
let generator: RandomNumberGenerator
class SomeSuperClass { init(sides: Int, generator: RandomNumberGenerator) {
init() { self.sides = sides
// initializer implementation goes here self.generator = generator
} }
} func roll() -> Int {
return Int(generator.random() * Double(sides)) + 1
class SomeSubClass: SomeSuperClass, SomeProtocol { }
// "required" from SomeProtocol conformance; "override" from }
SomeSuperClass
This example defines a new class called Dice, which represents an n-sided dice for
required override init() { use in a board game. Dice instances have an integer property called sides, which
// initializer implementation goes here represents how many sides they have, and a property called generator, which
} provides a random number generator from which to create dice roll values.
}
The generator property is of type RandomNumberGenerator. Therefore, you can set it
Failable Initializer Requirements to an instance of any type that adopts the RandomNumberGenerator protocol. Nothing
Protocols can define failable initializer requirements for conforming types, as defined else is required of the instance you assign to this property, except that the instance
in Failable Initializers. must adopt the RandomNumberGenerator protocol.

A failable initializer requirement can be satisfied by a failable or nonfailable initializer Dice also has an initializer, to set up its initial state. This initializer has a parameter
on a conforming type. A nonfailable initializer requirement can be satisfied by a called generator, which is also of type RandomNumberGenerator. You can pass a value
nonfailable initializer or an implicitly unwrapped failable initializer. of any conforming type in to this parameter when initializing a new Dice instance.

Dice provides one instance method, roll, which returns an integer value between 1
Protocols as Types
and the number of sides on the dice. This method calls the generator’s random()
Protocols do not actually implement any functionality themselves. Nonetheless, any method to create a new random number between 0.0 and 1.0, and uses this random
protocol you create will become a fully-fledged type for use in your code. number to create a dice roll value within the correct range. Because generator is
known to adopt RandomNumberGenerator, it is guaranteed to have a random() method
Because it is a type, you can use a protocol in many places where other types are to call.
allowed, including:
Here’s how the Dice class can be used to create a six-sided dice with a
As a parameter type or return type in a function, method, or initializer LinearCongruentialGenerator instance as its random number generator:
As the type of a constant, variable, or property
var d6 = Dice(sides: 6, generator: LinearCongruentialGenerator())
As the type of items in an array, dictionary, or other container
for _ in 1...5 {
NOTE

153
 print("Random dice roll is \(d6.roll())") var board: [Int]
} init() {
// Random dice roll is 3 board = Array(repeating: 0, count: finalSquare + 1)
// Random dice roll is 5 board[03] = +08; board[06] = +11; board[09] = +09;
board[10] = +02
// Random dice roll is 4
board[14] = -10; board[19] = -11; board[22] = -02;
// Random dice roll is 5
board[24] = -08
// Random dice roll is 4
}
var delegate: DiceGameDelegate?
Delegation
func play() {
Delegation is a design pattern that enables a class or structure to hand off (or
square = 0
delegate) some of its responsibilities to an instance of another type. This design
pattern is implemented by defining a protocol that encapsulates the delegated delegate?.gameDidStart(self)
responsibilities, such that a conforming type (known as a delegate) is guaranteed to gameLoop: while square != finalSquare {
provide the functionality that has been delegated. Delegation can be used to respond let diceRoll = dice.roll()
to a particular action, or to retrieve data from an external source without needing to
delegate?.game(self, didStartNewTurnWithDiceRoll:
know the underlying type of that source. diceRoll)
switch square + diceRoll {
The example below defines two protocols for use with dice-based board games:
case finalSquare:
protocol DiceGame { break gameLoop
var dice: Dice { get } case let newSquare where newSquare > finalSquare:
func play() continue gameLoop
} default:
protocol DiceGameDelegate { square += diceRoll
func gameDidStart(_ game: DiceGame) square += board[square]
func game(_ game: DiceGame, didStartNewTurnWithDiceRoll }
diceRoll: Int) }
func gameDidEnd(_ game: DiceGame) delegate?.gameDidEnd(self)
} }
The DiceGame protocol is a protocol that can be adopted by any game that involves }
dice. The DiceGameDelegate protocol can be adopted by any type to track the
For a description of the Snakes and Ladders gameplay, see Break section of the
progress of a DiceGame.
Control Flow.
Here’s a version of the Snakes and Ladders game originally introduced in Control
This version of the game is wrapped up as a class called SnakesAndLadders, which
Flow. This version is adapted to use a Dice instance for its dice-rolls; to adopt the
adopts the DiceGame protocol. It provides a gettable dice property and a play()
DiceGame protocol; and to notify a DiceGameDelegate about its progress:
method in order to conform to the protocol. (The dice property is declared as a
class SnakesAndLadders: DiceGame {
constant property because it does not need to change after initialization, and the
protocol only requires that it is gettable.)
let finalSquare = 25
let dice = Dice(sides: 6, generator: The Snakes and Ladders game board setup takes place within the class’s init()
LinearCongruentialGenerator())
initializer. All game logic is moved into the protocol’s play method, which uses the
var square = 0 protocol’s required dice property to provide its dice roll values.

154
Note that the delegate property is defined as an optional DiceGameDelegate, because The implementation of gameDidStart(_:) shown above uses the game parameter to
a delegate isn’t required in order to play the game. Because it is of an optional type, print some introductory information about the game that is about to be played. The
the delegate property is automatically set to an initial value of nil. Thereafter, the game parameter has a type of DiceGame, not SnakesAndLadders, and so
game instantiator has the option to set the property to a suitable delegate. gameDidStart(_:) can access and use only methods and properties that are
implemented as part of the DiceGame protocol. However, the method is still able to use
DiceGameDelegate provides three methods for tracking the progress of a game. These type casting to query the type of the underlying instance. In this example, it checks
three methods have been incorporated into the game logic within the play() method whether game is actually an instance of SnakesAndLadders behind the scenes, and
above, and are called when a new game starts, a new turn begins, or the game ends. prints an appropriate message if so.

Because the delegate property is an optional DiceGameDelegate, the play() method The gameDidStart(_:) method also accesses the dice property of the passed game
uses optional chaining each time it calls a method on the delegate. If the delegate parameter. Because game is known to conform to the DiceGame protocol, it is
property is nil, these delegate calls fail gracefully and without error. If the delegate guaranteed to have a dice property, and so the gameDidStart(_:) method is able to
property is non-nil, the delegate methods are called, and are passed the access and print the dice’s sides property, regardless of what kind of game is being
SnakesAndLadders instance as a parameter. played.

This next example shows a class called DiceGameTracker, which adopts the Here’s how DiceGameTracker looks in action:
DiceGameDelegate protocol:
let tracker = DiceGameTracker()
class DiceGameTracker: DiceGameDelegate { let game = SnakesAndLadders()
var numberOfTurns = 0 game.delegate = tracker
func gameDidStart(_ game: DiceGame) { game.play()
numberOfTurns = 0 // Started a new game of Snakes and Ladders
if game is SnakesAndLadders { // The game is using a 6-sided dice
print("Started a new game of Snakes and Ladders") // Rolled a 3
} // Rolled a 5
print("The game is using a \(game.dice.sides)-sided dice") // Rolled a 4
} // Rolled a 5
func game(_ game: DiceGame, didStartNewTurnWithDiceRoll // The game lasted for 4 turns
diceRoll: Int) {
numberOfTurns += 1 Adding Protocol Conformance with an Extension
print("Rolled a \(diceRoll)")
You can extend an existing type to adopt and conform to a new protocol, even if you
} do not have access to the source code for the existing type. Extensions can add new
func gameDidEnd(_ game: DiceGame) { properties, methods, and subscripts to an existing type, and are therefore able to add
print("The game lasted for \(numberOfTurns) turns") any requirements that a protocol may demand. For more about extensions, see
Extensions.
}
} NOTE
DiceGameTracker implements all three methods required by DiceGameDelegate. It Existing instances of a type automatically adopt and conform to a protocol when that
uses these methods to keep track of the number of turns a game has taken. It resets a conformance is added to the instance’s type in an extension.
numberOfTurns property to zero when the game starts, increments it each time a new
turn begins, and prints out the total number of turns once the game has ended. For example, this protocol, called TextRepresentable, can be implemented by any
type that has a way to be represented as text. This might be a description of itself, or a
text version of its current state:

155
 protocol TextRepresentable { return "A hamster named \(name)"
var textualDescription: String { get } }
} }

The Dice class from earlier can be extended to adopt and conform to extension Hamster: TextRepresentable {}
TextRepresentable: Instances of Hamster can now be used wherever TextRepresentable is the required
type:
extension Dice: TextRepresentable {
var textualDescription: String { let simonTheHamster = Hamster(name: "Simon")
return "A \(sides)-sided dice" let somethingTextRepresentable: TextRepresentable = simonTheHamster
} print(somethingTextRepresentable.textualDescription)
} // Prints "A hamster named Simon"

This extension adopts the new protocol in exactly the same way as if Dice had NOTE
provided it in its original implementation. The protocol name is provided after the type Types do not automatically adopt a protocol just by satisfying its requirements. They
name, separated by a colon, and an implementation of all requirements of the protocol must always explicitly declare their adoption of the protocol.
is provided within the extension’s curly braces.
Collections of Protocol Types
Any Dice instance can now be treated as TextRepresentable:
A protocol can be used as the type to be stored in a collection such as an array or a
dictionary, as mentioned in Protocols as Types. This example creates an array of
let d12 = Dice(sides: 12, generator: LinearCongruentialGenerator())
TextRepresentable things:
print(d12.textualDescription)
// Prints "A 12-sided dice" let things: [TextRepresentable] = [game, d12, simonTheHamster]

Similarly, the SnakesAndLadders game class can be extended to adopt and conform to It is now possible to iterate over the items in the array, and print each item’s textual
the TextRepresentable protocol: description:

extension SnakesAndLadders: TextRepresentable { for thing in things {

var textualDescription: String { print(thing.textualDescription)
return "A game of Snakes and Ladders with \(finalSquare) }
squares"
// A game of Snakes and Ladders with 25 squares
}
// A 12-sided dice
}
// A hamster named Simon
print(game.textualDescription)
Note that the thing constant is of type TextRepresentable. It is not of type Dice, or
// Prints "A game of Snakes and Ladders with 25 squares"
DiceGame, or Hamster, even if the actual instance behind the scenes is of one of those

Declaring Protocol Adoption with an Extension types. Nonetheless, because it is of type TextRepresentable, and anything that is
TextRepresentable is known to have a textualDescription property, it is safe to
If a type already conforms to all of the requirements of a protocol, but has not yet access thing.textualDescription each time through the loop.
stated that it adopts that protocol, you can make it adopt the protocol with an empty
extension:
Protocol Inheritance
struct Hamster { A protocol can inherit one or more other protocols and can add further requirements
var name: String on top of the requirements it inherits. The syntax for protocol inheritance is similar to
the syntax for class inheritance, but with the option to list multiple inherited protocols,
var textualDescription: String {
separated by commas:

156
 protocol InheritingProtocol: SomeProtocol, AnotherProtocol { start of its pretty text representation. It then iterates through the array of board
// protocol definition goes here squares, and appends a geometric shape to represent the contents of each square:
}
If the square’s value is greater than 0, it is the base of a ladder, and is
Here’s an example of a protocol that inherits the TextRepresentable protocol from
represented by ▲.
above:
If the square’s value is less than 0, it is the head of a snake, and is
protocol PrettyTextRepresentable: TextRepresentable { represented by ▼.
var prettyTextualDescription: String { get } Otherwise, the square’s value is 0, and it is a “free” square, represented by
} ○.

This example defines a new protocol, PrettyTextRepresentable, which inherits from The prettyTextualDescription property can now be used to print a pretty text
TextRepresentable. Anything that adopts PrettyTextRepresentable must satisfy all description of any SnakesAndLadders instance:
of the requirements enforced by TextRepresentable, plus the additional requirements
enforced by PrettyTextRepresentable. In this example, PrettyTextRepresentable print(game.prettyTextualDescription)
adds a single requirement to provide a gettable property called // A game of Snakes and Ladders with 25 squares:
prettyTextualDescription that returns a String. // ○ ○ ▲ ○ ○ ▲ ○ ○ ▲ ▲ ○ ○ ○ ▼ ○ ○ ○ ○ ▼ ○ ○ ▼ ○ ▼ ○

The SnakesAndLadders class can be extended to adopt and conform to

Class-Only Protocols
PrettyTextRepresentable:
You can limit protocol adoption to class types (and not structures or enumerations) by
extension SnakesAndLadders: PrettyTextRepresentable { adding the class keyword to a protocol’s inheritance list. The class keyword must
always appear first in a protocol’s inheritance list, before any inherited protocols:
var prettyTextualDescription: String {
var output = textualDescription + ":\n" protocol SomeClassOnlyProtocol: class, SomeInheritedProtocol {
for index in 1...finalSquare { // class-only protocol definition goes here
switch board[index] { }
case let ladder where ladder > 0:
In the example above, SomeClassOnlyProtocol can only be adopted by class types. It
output += "▲ " is a compile-time error to write a structure or enumeration definition that tries to adopt
case let snake where snake < 0: SomeClassOnlyProtocol.
output += "▼ "
NOTE
default:
output += "○ " Use a class-only protocol when the behavior defined by that protocol’s requirements
assumes or requires that a conforming type has reference semantics rather than value
} semantics. For more on reference and value semantics, see Structures and
} Enumerations Are Value Types and Classes Are Reference Types.
return output
}
Protocol Composition
}
It can be useful to require a type to conform to multiple protocols at once. You can
combine multiple protocols into a single requirement with a protocol composition.
This extension states that it adopts the PrettyTextRepresentable protocol and Protocol compositions have the form SomeProtocol & AnotherProtocol. You can list
provides an implementation of the prettyTextualDescription property for the as many protocols as you need to, separating them by ampersands (&).
SnakesAndLadders type. Anything that is PrettyTextRepresentable must also be
TextRepresentable, and so the implementation of prettyTextualDescription starts Here’s an example that combines two protocols called Named and Aged into a single
by accessing the textualDescription property from the TextRepresentable protocol protocol composition requirement on a function parameter:
to begin an output string. It appends a colon and a line break, and uses this as the

157
 protocol Named { The is operator returns true if an instance conforms to a protocol and
var name: String { get } returns false if it does not.
} The as? version of the downcast operator returns an optional value of the
protocol Aged { protocol’s type, and this value is nil if the instance does not conform to
var age: Int { get } that protocol.
} The as! version of the downcast operator forces the downcast to the
struct Person: Named, Aged { protocol type and triggers a runtime error if the downcast does not
succeed.
var name: String
var age: Int This example defines a protocol called HasArea, with a single property requirement of
a gettable Double property called area:
}
func wishHappyBirthday(to celebrator: Named & Aged) { protocol HasArea {
print("Happy birthday, \(celebrator.name), you're \ var area: Double { get }
(celebrator.age)!")
}
}
let birthdayPerson = Person(name: "Malcolm", age: 21)
Here are two classes, Circle and Country, both of which conform to the HasArea
protocol:
wishHappyBirthday(to: birthdayPerson)
// Prints "Happy birthday, Malcolm, you're 21!" class Circle: HasArea {
This example defines a protocol called Named, with a single requirement for a gettable let pi = 3.1415927
String property called name. It also defines a protocol called Aged, with a single var radius: Double
requirement for a gettable Int property called age. Both of these protocols are var area: Double { return pi * radius * radius }
adopted by a structure called Person.
init(radius: Double) { self.radius = radius }

The example also defines a wishHappyBirthday(to:) function, The type of the }

celebrator parameter is Named & Aged, which means “any type that conforms to both class Country: HasArea {
the Named and Aged protocols.” It doesn’t matter what specific type is passed to the var area: Double
function, as long as it conforms to both of the required protocols. init(area: Double) { self.area = area }
}
The example then creates a new Person instance called birthdayPerson and passes
this new instance to the wishHappyBirthday(to:) function. Because Person conforms The Circle class implements the area property requirement as a computed property,
to both protocols, this is a valid call, and the wishHappyBirthday(to:) function is able based on a stored radius property. The Country class implements the area
to print its birthday greeting. requirement directly as a stored property. Both classes correctly conform to the
HasArea protocol.
NOTE

Protocol compositions do not define a new, permanent protocol type. Rather, they
Here’s a class called Animal, which does not conform to the HasArea protocol:
define a temporary local protocol that has the combined requirements of all protocols in
the composition. class Animal {
var legs: Int
Checking for Protocol Conformance init(legs: Int) { self.legs = legs }
You can use the is and as operators described in Type Casting to check for protocol }
conformance, and to cast to a specific protocol. Checking for and casting to a protocol
follows exactly the same syntax as checking for and casting to a type:

158
The Circle, Country and Animal classes do not have a shared base class. @objc attribute. Note that @objc protocols can be adopted only by classes that inherit
Nonetheless, they are all classes, and so instances of all three types can be used to from Objective-C classes or other @objc classes. They can’t be adopted by structures
initialize an array that stores values of type AnyObject: or enumerations.

let objects: [AnyObject] = [ When you use a method or property in an optional requirement, its type automatically
Circle(radius: 2.0), becomes an optional. For example, a method of type (Int) -> String becomes
Country(area: 243_610), ((Int) -> String)?. Note that the entire function type is wrapped in the optional, not
the method’s return value.
Animal(legs: 4)
]
An optional protocol requirement can be called with optional chaining, to account for
The objects array is initialized with an array literal containing a Circle instance with a the possibility that the requirement was not implemented by a type that conforms to
radius of 2 units; a Country instance initialized with the surface area of the United the protocol. You check for an implementation of an optional method by writing a
Kingdom in square kilometers; and an Animal instance with four legs. question mark after the name of the method when it is called, such as
someOptionalMethod?(someArgument). For information on optional chaining, see
The objects array can now be iterated, and each object in the array can be checked Optional Chaining.
to see if it conforms to the HasArea protocol:
The following example defines an integer-counting class called Counter, which uses
for object in objects { an external data source to provide its increment amount. This data source is defined
if let objectWithArea = object as? HasArea { by the CounterDataSource protocol, which has two optional requirements:
print("Area is \(objectWithArea.area)")
@objc protocol CounterDataSource {
} else {
@objc optional func increment(forCount count: Int) -> Int
print("Something that doesn't have an area")
@objc optional var fixedIncrement: Int { get }
}
}
}
The CounterDataSource protocol defines an optional method requirement called
// Area is 12.5663708
increment(forCount:) and an optional property requirement called fixedIncrement.
// Area is 243610.0 These requirements define two different ways for data sources to provide an
// Something that doesn't have an area appropriate increment amount for a Counter instance.
Whenever an object in the array conforms to the HasArea protocol, the optional value
NOTE
returned by the as? operator is unwrapped with optional binding into a constant called
objectWithArea. The objectWithArea constant is known to be of type HasArea, and Strictly speaking, you can write a custom class that conforms to CounterDataSource
so its area property can be accessed and printed in a type-safe way. without implementing either protocol requirement. They are both optional, after all.
Although technically allowed, this wouldn’t make for a very good data source.
Note that the underlying objects are not changed by the casting process. They The Counter class, defined below, has an optional dataSource property of type
continue to be a Circle, a Country and an Animal. However, at the point that they are CounterDataSource?:
stored in the objectWithArea constant, they are only known to be of type HasArea,
and so only their area property can be accessed. class Counter {
var count = 0
Optional Protocol Requirements var dataSource: CounterDataSource?
You can define optional requirements for protocols, These requirements do not have func increment() {
to be implemented by types that conform to the protocol. Optional requirements are
if let amount = dataSource?.increment?(forCount: count) {
prefixed by the optional modifier as part of the protocol’s definition. Optional
requirements are available so that you can write code that interoperates with count += amount
Objective-C. Both the protocol and the optional requirement must be marked with the } else if let amount = dataSource?.fixedIncrement {

159
 count += amount Here’s a simple CounterDataSource implementation where the data source returns a
} constant value of 3 every time it is queried. It does this by implementing the optional
} fixedIncrement property requirement:

}
class ThreeSource: NSObject, CounterDataSource {
The Counter class stores its current value in a variable property called count. The let fixedIncrement = 3
Counter class also defines a method called increment, which increments the count
}
property every time the method is called.
You can use an instance of ThreeSource as the data source for a new Counter
The increment() method first tries to retrieve an increment amount by looking for an instance:
implementation of the increment(forCount:) method on its data source. The
increment() method uses optional chaining to try to call increment(forCount:), and var counter = Counter()
passes the current count value as the method’s single argument. counter.dataSource = ThreeSource()
for _ in 1...4 {
Note that two levels of optional chaining are at play here. First, it is possible that counter.increment()
dataSource may be nil, and so dataSource has a question mark after its name to
print(counter.count)
indicate that increment(forCount:) should be called only if dataSource isn’t nil.
Second, even if dataSource does exist, there is no guarantee that it implements }
increment(forCount:), because it is an optional requirement. Here, the possibility // 3
that increment(forCount:) might not be implemented is also handled by optional // 6
chaining. The call to increment(forCount:) happens only if increment(forCount:) // 9
exists—that is, if it isn’t nil. This is why increment(forCount:) is also written with a
// 12
question mark after its name.
The code above creates a new Counter instance; sets its data source to be a new
Because the call to increment(forCount:) can fail for either of these two reasons, the ThreeSource instance; and calls the counter’s increment() method four times. As
call returns an optional Int value. This is true even though increment(forCount:) is expected, the counter’s count property increases by three each time increment() is
defined as returning a nonoptional Int value in the definition of CounterDataSource. called.
Even though there are two optional chaining operations, one after another, the result
is still wrapped in a single optional. For more information about using multiple optional Here’s a more complex data source called TowardsZeroSource, which makes a
chaining operations, see Linking Multiple Levels of Chaining. Counter instance count up or down towards zero from its current count value:

After calling increment(forCount:), the optional Int that it returns is unwrapped into @objc class TowardsZeroSource: NSObject, CounterDataSource {
a constant called amount, using optional binding. If the optional Int does contain a func increment(forCount count: Int) -> Int {
value—that is, if the delegate and method both exist, and the method returned a value if count == 0 {
—the unwrapped amount is added onto the stored count property, and incrementation return 0
is complete.
} else if count < 0 {

If it is not possible to retrieve a value from the increment(forCount:) method—either return 1

because dataSource is nil, or because the data source does not implement } else {
increment(forCount:)—then the increment() method tries to retrieve a value from return -1
the data source’s fixedIncrement property instead. The fixedIncrement property is }
also an optional requirement, so its value is an optional Int value, even though
}
fixedIncrement is defined as a nonoptional Int property as part of the
CounterDataSource protocol definition. }

160
The TowardsZeroSource class implements the optional increment(forCount:) // Prints "And here's a random Boolean: true"
method from the CounterDataSource protocol and uses the count argument value to
work out which direction to count in. If count is already zero, the method returns 0 to Providing Default Implementations
indicate that no further counting should take place. You can use protocol extensions to provide a default implementation to any method or
computed property requirement of that protocol. If a conforming type provides its own
You can use an instance of TowardsZeroSource with the existing Counter instance to implementation of a required method or property, that implementation will be used
count from -4 to zero. Once the counter reaches zero, no more counting takes place: instead of the one provided by the extension.

counter.count = -4 NOTE

counter.dataSource = TowardsZeroSource() Protocol requirements with default implementations provided by extensions are distinct
for _ in 1...5 { from optional protocol requirements. Although conforming types don’t have to provide
their own implementation of either, requirements with default implementations can be
counter.increment() called without optional chaining.
print(counter.count)
For example, the PrettyTextRepresentable protocol, which inherits the
} TextRepresentable protocol can provide a default implementation of its required
// -3 prettyTextualDescription property to simply return the result of accessing the
// -2 textualDescription property:
// -1
extension PrettyTextRepresentable {
// 0
var prettyTextualDescription: String {
// 0
return textualDescription
}
Protocol Extensions
}
Protocols can be extended to provide method and property implementations to
conforming types. This allows you to define behavior on protocols themselves, rather Adding Constraints to Protocol Extensions
than in each type’s individual conformance or in a global function.
When you define a protocol extension, you can specify constraints that conforming
types must satisfy before the methods and properties of the extension are available.
For example, the RandomNumberGenerator protocol can be extended to provide a
You write these constraints after the name of the protocol you’re extending using a
randomBool() method, which uses the result of the required random() method to
generic where clause, as described in Generic Where Clauses.
return a random Bool value:

extension RandomNumberGenerator { For instance, you can define an extension to the Collection protocol that applies to
any collection whose elements conform to the TextRepresentable protocol from the
func randomBool() -> Bool {
example above.
return random() > 0.5
} extension Collection where Iterator.Element: TextRepresentable {
} var textualDescription: String {

By creating an extension on the protocol, all conforming types automatically gain this let itemsAsText = self.map { $0.textualDescription }
method implementation without any additional modification. return "[" + itemsAsText.joined(separator: ", ") + "]"
}
let generator = LinearCongruentialGenerator()
}
print("Here's a random number: \(generator.random())")
The textualDescription property returns the textual description of the entire
// Prints "Here's a random number: 0.37464991998171"
collection by concatenating the textual representation of each element in the collection
print("And here's a random Boolean: \(generator.randomBool())") into a comma-separated list, enclosed in brackets.

161
Consider the Hamster structure from before, which conforms to the
TextRepresentable protocol, and an array of Hamster values:

let murrayTheHamster = Hamster(name: "Murray")

let morganTheHamster = Hamster(name: "Morgan")
let mauriceTheHamster = Hamster(name: "Maurice")
let hamsters = [murrayTheHamster, morganTheHamster,
mauriceTheHamster]

Because Array conforms to Collection and the array’s elements conform to the
TextRepresentable protocol, the array can use the textualDescription property to
get a textual representation of its contents:

print(hamsters.textualDescription)
// Prints "[A hamster named Murray, A hamster named Morgan, A
hamster named Maurice]"
NOTE

If a conforming type satisfies the requirements for multiple constrained extensions that
provide implementations for the same method or property, Swift will use the
implementation corresponding to the most specialized constraints.

162
 Section 23 The swapTwoInts(_:_:) function is useful, but it can only be used with Int values. If
you want to swap two String values, or two Double values, you have to write more

Generics functions, such as the swapTwoStrings(_:_:) and swapTwoDoubles(_:_:) functions

shown below:

func swapTwoStrings(_ a: inout String, _ b: inout String) {

let temporaryA = a
Generics a = b
b = temporaryA
Generic code enables you to write flexible, reusable functions and types that can work }
with any type, subject to requirements that you define. You can write code that avoids
duplication and expresses its intent in a clear, abstracted manner.
func swapTwoDoubles(_ a: inout Double, _ b: inout Double) {
Generics are one of the most powerful features of Swift, and much of the Swift let temporaryA = a
standard library is built with generic code. In fact, you’ve been using generics a = b
throughout the Language Guide, even if you didn’t realize it. For example, Swift’s b = temporaryA
Array and Dictionary types are both generic collections. You can create an array that
}
holds Int values, or an array that holds String values, or indeed an array for any
other type that can be created in Swift. Similarly, you can create a dictionary to store You may have noticed that the bodies of the swapTwoInts(_:_:),
values of any specified type, and there are no limitations on what that type can be. swapTwoStrings(_:_:), and swapTwoDoubles(_:_:) functions are identical. The only
difference is the type of the values that they accept (Int, String, and Double).
The Problem That Generics Solve
It would be much more useful, and considerably more flexible, to write a single
Here’s a standard, non-generic function called swapTwoInts(_:_:), which swaps two
function that could swap two values of any type. Generic code enables you to write
Int values:
such a function. (A generic version of these functions is defined below.)
func swapTwoInts(_ a: inout Int, _ b: inout Int) {
NOTE
let temporaryA = a
In all three functions, it is important that the types of a and b are defined to be the same
a = b
as each other. If a and b were not of the same type, it would not be possible to swap
b = temporaryA their values. Swift is a type-safe language, and does not allow (for example) a variable
} of type String and a variable of type Double to swap values with each other.
Attempting to do so would be reported as a compile-time error.
This function makes use of in-out parameters to swap the values of a and b, as
described in In-Out Parameters. Generic Functions
Generic functions can work with any type. Here’s a generic version of the
The swapTwoInts(_:_:) function swaps the original value of b into a, and the original
swapTwoInts(_:_:) function from above, called swapTwoValues(_:_:):
value of a into b. You can call this function to swap the values in two Int variables:
func swapTwoValues<T>(_ a: inout T, _ b: inout T) {
var someInt = 3
let temporaryA = a
var anotherInt = 107
a = b
swapTwoInts(&someInt, &anotherInt)
b = temporaryA
print("someInt is now \(someInt), and anotherInt is now \
(anotherInt)") }
// Prints "someInt is now 107, and anotherInt is now 3"

163
The body of the swapTwoValues(_:_:) function is identical to the body of the In the swapTwoValues(_:_:) example above, the placeholder type T is an example of
swapTwoInts(_:_:) function. However, the first line of swapTwoValues(_:_:) is slightly a type parameter. Type parameters specify and name a placeholder type, and are
different from swapTwoInts(_:_:). Here’s how the first lines compare: written immediately after the function’s name, between a pair of matching angle
brackets (such as <T>).
func swapTwoInts(_ a: inout Int, _ b: inout Int)
func swapTwoValues<T>(_ a: inout T, _ b: inout T) Once you specify a type parameter, you can use it to define the type of a function’s
parameters (such as the a and b parameters of the swapTwoValues(_:_:) function), or
The generic version of the function uses a placeholder type name (called T, in this
as the function’s return type, or as a type annotation within the body of the function. In
case) instead of an actual type name (such as Int, String, or Double). The
each case, the type parameter is replaced with an actual type whenever the function
placeholder type name doesn’t say anything about what T must be, but it does say
is called. (In the swapTwoValues(_:_:) example above, T was replaced with Int the
that both a and b must be of the same type T, whatever T represents. The actual type
first time the function was called, and was replaced with String the second time it was
to use in place of T will be determined each time the swapTwoValues(_:_:) function is
called.)
called.

You can provide more than one type parameter by writing multiple type parameter
The other difference is that the generic function’s name (swapTwoValues(_:_:)) is
names within the angle brackets, separated by commas.
followed by the placeholder type name (T) inside angle brackets (<T>). The brackets
tell Swift that T is a placeholder type name within the swapTwoValues(_:_:) function
definition. Because T is a placeholder, Swift does not look for an actual type called T. Naming Type Parameters
In most cases, type parameters have descriptive names, such as Key and Value in
The swapTwoValues(_:_:) function can now be called in the same way as Dictionary<Key, Value> and Element in Array<Element>, which tells the reader
swapTwoInts, except that it can be passed two values of any type, as long as both of about the relationship between the type parameter and the generic type or function it’s
those values are of the same type as each other. Each time swapTwoValues(_:_:) is used in. However, when there isn’t a meaningful relationship between them, it’s
called, the type to use for T is inferred from the types of values passed to the function. traditional to name them using single letters such as T, U, and V, such as T in the
swapTwoValues(_:_:) function above.
In the two examples below, T is inferred to be Int and String respectively:
NOTE
var someInt = 3
Always give type parameters upper camel case names (such as T and
var anotherInt = 107 MyTypeParameter) to indicate that they are a placeholder for a type, not a value.
swapTwoValues(&someInt, &anotherInt)
// someInt is now 107, and anotherInt is now 3 Generic Types
In addition to generic functions, Swift enables you to define your own generic types.
These are custom classes, structures, and enumerations that can work with any type,
var someString = "hello"
in a similar way to Array and Dictionary.
var anotherString = "world"
swapTwoValues(&someString, &anotherString) This section shows you how to write a generic collection type called Stack. A stack is
// someString is now "world", and anotherString is now "hello" an ordered set of values, similar to an array, but with a more restricted set of
NOTE operations than Swift’s Array type. An array allows new items to be inserted and
removed at any location in the array. A stack, however, allows new items to be
The swapTwoValues(_:_:) function defined above is inspired by a generic function
appended only to the end of the collection (known as pushing a new value on to the
called swap, which is part of the Swift standard library, and is automatically made
available for you to use in your apps. If you need the behavior of the stack). Similarly, a stack allows items to be removed only from the end of the
swapTwoValues(_:_:) function in your own code, you can use Swift’s existing collection (known as popping a value off the stack).
swap(_:_:) function rather than providing your own implementation.
NOTE

Type Parameters The concept of a stack is used by the UINavigationController class to model the
view controllers in its navigation hierarchy. You call the UINavigationController class

164
 pushViewController(_:animated:) method to add (or push) a view controller on to Here’s a generic version of the same code:
the navigation stack, and its popViewControllerAnimated(_:) method to remove (or
pop) a view controller from the navigation stack. A stack is a useful collection model struct Stack<Element> {
whenever you need a strict “last in, first out” approach to managing a collection.
var items = [Element]()
The illustration below shows the push / pop behavior for a stack: mutating func push(_ item: Element) {
items.append(item)
}
mutating func pop() -> Element {
return items.removeLast()
}
}

Note how the generic version of Stack is essentially the same as the non-generic
version, but with a type parameter called Element instead of an actual type of Int.
This type parameter is written within a pair of angle brackets (<Element>) immediately
1.There are currently three values on the stack. after the structure’s name.

2.A fourth value is “pushed” on to the top of the stack. Element defines a placeholder name for “some type Element” to be provided later on.
3.The stack now holds four values, with the most recent one at the top. This future type can be referred to as “Element” anywhere within the structure’s
definition. In this case, Element is used as a placeholder in three places:
4.The top item in the stack is removed, or “popped”.
5.After popping a value, the stack once again holds three values. To create a property called items, which is initialized with an empty array of
Here’s how to write a non-generic version of a stack, in this case for a stack of Int values of type Element
values: To specify that the push(_:) method has a single parameter called item,
which must be of type Element
struct IntStack {
To specify that the value returned by the pop() method will be a value of
var items = [Int]()
type Element
mutating func push(_ item: Int) {
Because it is a generic type, Stack can be used to create a stack of any valid type in
items.append(item)
Swift, in a similar manner to Array and Dictionary.
}
mutating func pop() -> Int { You create a new Stack instance by writing the type to be stored in the stack within
return items.removeLast() angle brackets. For example, to create a new stack of strings, you write
} Stack<String>():

}
var stackOfStrings = Stack<String>()
This structure uses an Array property called items to store the values in the stack. stackOfStrings.push("uno")
Stack provides two methods, push and pop, to push and pop values on and off the
stackOfStrings.push("dos")
stack. These methods are marked as mutating, because they need to modify (or
stackOfStrings.push("tres")
mutate) the structure’s items array.
stackOfStrings.push("cuatro")
The IntStack type shown above can only be used with Int values, however. It would // the stack now contains 4 strings
be much more useful to define a generic Stack class, that can manage a stack of any Here’s how stackOfStrings looks after pushing these four values on to the stack:
type of value.

165
 The topItem property returns an optional value of type Element. If the stack is empty,
topItem returns nil; if the stack is not empty, topItem returns the final item in the
items array.

Note that this extension does not define a type parameter list. Instead, the Stack
type’s existing type parameter name, Element, is used within the extension to indicate
the optional type of the topItem computed property.

Popping a value from the stack removes and returns the top value, "cuatro": The topItem computed property can now be used with any Stack instance to access
and query its top item without removing it:
let fromTheTop = stackOfStrings.pop()
// fromTheTop is equal to "cuatro", and the stack now contains 3 if let topItem = stackOfStrings.topItem {
strings
print("The top item on the stack is \(topItem).")
Here’s how the stack looks after popping its top value: }
// Prints "The top item on the stack is tres."

Type Constraints
The swapTwoValues(_:_:) function and the Stack type can work with any type.
However, it is sometimes useful to enforce certain type constraints on the types that
can be used with generic functions and generic types. Type constraints specify that a
type parameter must inherit from a specific class, or conform to a particular protocol or
protocol composition.

For example, Swift’s Dictionary type places a limitation on the types that can be used
as keys for a dictionary. As described in Dictionaries, the type of a dictionary’s keys
must be hashable. That is, it must provide a way to make itself uniquely
Extending a Generic Type representable. Dictionary needs its keys to be hashable so that it can check whether
it already contains a value for a particular key. Without this requirement, Dictionary
When you extend a generic type, you do not provide a type parameter list as part of could not tell whether it should insert or replace a value for a particular key, nor would
the extension’s definition. Instead, the type parameter list from the original type it be able to find a value for a given key that is already in the dictionary.
definition is available within the body of the extension, and the original type parameter
names are used to refer to the type parameters from the original definition.
This requirement is enforced by a type constraint on the key type for Dictionary,
which specifies that the key type must conform to the Hashable protocol, a special
The following example extends the generic Stack type to add a read-only computed protocol defined in the Swift standard library. All of Swift’s basic types (such as
property called topItem, which returns the top item on the stack without popping it String, Int, Double, and Bool) are hashable by default.
from the stack:
You can define your own type constraints when creating custom generic types, and
extension Stack {
these constraints provide much of the power of generic programming. Abstract
var topItem: Element? {
concepts like Hashable characterize types in terms of their conceptual characteristics,
return items.isEmpty ? nil : items[items.count - 1] rather than their explicit type.
}
} Type Constraint Syntax
You write type constraints by placing a single class or protocol constraint after a type
parameter’s name, separated by a colon, as part of the type parameter list. The basic

166
syntax for type constraints on a generic function is shown below (although the syntax because the function returns an optional index number, not an optional value from the
is the same for generic types): array. Be warned, though—this function does not compile, for reasons explained after
the example:
func someFunction<T: SomeClass, U: SomeProtocol>(someT: T, someU:
U) { func findIndex<T>(of valueToFind: T, in array:[T]) -> Int? {
// function body goes here for (index, value) in array.enumerated() {
} if value == valueToFind {
The hypothetical function above has two type parameters. The first type parameter, T, return index
has a type constraint that requires T to be a subclass of SomeClass. The second type }
parameter, U, has a type constraint that requires U to conform to the protocol
}
SomeProtocol.
return nil

Type Constraints in Action }

Here’s a non-generic function called findIndex(ofString:in:), which is given a This function does not compile as written above. The problem lies with the equality
String value to find and an array of String values within which to find it. The check, “if value == valueToFind”. Not every type in Swift can be compared with the
findIndex(ofString:in:) function returns an optional Int value, which will be the equal to operator (==). If you create your own class or structure to represent a
index of the first matching string in the array if it is found, or nil if the string cannot be complex data model, for example, then the meaning of “equal to” for that class or
found: structure is not something that Swift can guess for you. Because of this, it is not
possible to guarantee that this code will work for every possible type T, and an
func findIndex(ofString valueToFind: String, in array: [String]) -> appropriate error is reported when you try to compile the code.
Int? {
for (index, value) in array.enumerated() { All is not lost, however. The Swift standard library defines a protocol called Equatable,
if value == valueToFind { which requires any conforming type to implement the equal to operator (==) and the
return index not equal to operator (!=) to compare any two values of that type. All of Swift’s
standard types automatically support the Equatable protocol.
}
}
Any type that is Equatable can be used safely with the findIndex(of:in:) function,
return nil because it is guaranteed to support the equal to operator. To express this fact, you
} write a type constraint of Equatable as part of the type parameter’s definition when
you define the function:
The findIndex(ofString:in:) function can be used to find a string value in an array
of strings:
func findIndex<T: Equatable>(of valueToFind: T, in array:[T]) ->
Int? {
let strings = ["cat", "dog", "llama", "parakeet", "terrapin"]
for (index, value) in array.enumerated() {
if let foundIndex = findIndex(ofString: "llama", in: strings) {
if value == valueToFind {
print("The index of llama is \(foundIndex)")
return index
}
}
// Prints "The index of llama is 2"
}
The principle of finding the index of a value in an array isn’t useful only for strings, return nil
however. You can write the same functionality as a generic function by replacing any
}
mention of strings with values of some type T instead.
The single type parameter for findIndex(of:in:) is written as T: Equatable, which
Here’s how you might expect a generic version of findIndex(ofString:in:), called means “any type T that conforms to the Equatable protocol.”
findIndex(of:in:), to be written. Note that the return type of this function is still Int?,

167
The findIndex(of:in:) function now compiles successfully and can be used with any Any type that conforms to the Container protocol must be able to specify the type of
type that is Equatable, such as Double or String: values it stores. Specifically, it must ensure that only items of the right type are added
to the container, and it must be clear about the type of the items returned by its
let doubleIndex = findIndex(of: 9.3, in: [3.14159, 0.1, 0.25]) subscript.
// doubleIndex is an optional Int with no value, because 9.3 is not
in the array To define these requirements, the Container protocol needs a way to refer to the type
let stringIndex = findIndex(of: "Andrea", in: ["Mike", "Malcolm", of the elements that a container will hold, without knowing what that type is for a
"Andrea"]) specific container. The Container protocol needs to specify that any value passed to
// stringIndex is an optional Int containing a value of 2 the append(_:) method must have the same type as the container’s element type, and
that the value returned by the container’s subscript will be of the same type as the
Associated Types container’s element type.
When defining a protocol, it is sometimes useful to declare one or more associated
types as part of the protocol’s definition. An associated type gives a placeholder name To achieve this, the Container protocol declares an associated type called ItemType,
to a type that is used as part of the protocol. The actual type to use for that associated written as associatedtype ItemType. The protocol does not define what ItemType is
type is not specified until the protocol is adopted. Associated types are specified with —that information is left for any conforming type to provide. Nonetheless, the
ItemType alias provides a way to refer to the type of the items in a Container, and to
the associatedtype keyword.
define a type for use with the append(_:) method and subscript, to ensure that the
expected behavior of any Container is enforced.
Associated Types in Action
Here’s an example of a protocol called Container, which declares an associated type Here’s a version of the non-generic IntStack type from earlier, adapted to conform to
called ItemType: the Container protocol:
protocol Container {
struct IntStack: Container {
associatedtype ItemType
// original IntStack implementation
mutating func append(_ item: ItemType)
var items = [Int]()
var count: Int { get }
mutating func push(_ item: Int) {
subscript(i: Int) -> ItemType { get }
items.append(item)
}
}
The Container protocol defines three required capabilities that any container must mutating func pop() -> Int {
provide:
return items.removeLast()
}
It must be possible to add a new item to the container with an append(_:)
method. // conformance to the Container protocol
typealias ItemType = Int
It must be possible to access a count of the items in the container through
a count property that returns an Int value. mutating func append(_ item: Int) {
self.push(item)
It must be possible to retrieve each item in the container with a subscript
that takes an Int index value. }
var count: Int {
This protocol doesn’t specify how the items in the container should be stored or what
type they are allowed to be. The protocol only specifies the three bits of functionality return items.count
that any type must provide in order to be considered a Container. A conforming type }
can provide additional functionality, as long as it satisfies these three requirements. subscript(i: Int) -> Int {
return items[i]

168
 } This time, the type parameter Element is used as the type of the append(_:) method’s
} item parameter and the return type of the subscript. Swift can therefore infer that
Element is the appropriate type to use as the ItemType for this particular container.
The IntStack type implements all three of the Container protocol’s requirements, and
in each case wraps part of the IntStack type’s existing functionality to satisfy these
requirements. Extending an Existing Type to Specify an Associated Type
You can extend an existing type to add conformance to a protocol, as described in
Moreover, IntStack specifies that for this implementation of Container, the Adding Protocol Conformance with an Extension. This includes a protocol with an
appropriate ItemType to use is a type of Int. The definition of typealias ItemType = associated type.
Int turns the abstract type of ItemType into a concrete type of Int for this
implementation of the Container protocol. Swift’s Array type already provides an append(_:) method, a count property, and a
subscript with an Int index to retrieve its elements. These three capabilities match the
Thanks to Swift’s type inference, you don’t actually need to declare a concrete requirements of the Container protocol. This means that you can extend Array to
ItemType of Int as part of the definition of IntStack. Because IntStack conforms to conform to the Container protocol simply by declaring that Array adopts the protocol.
all of the requirements of the Container protocol, Swift can infer the appropriate You do this with an empty extension, as described in Declaring Protocol Adoption with
ItemType to use, simply by looking at the type of the append(_:) method’s item an Extension:
parameter and the return type of the subscript. Indeed, if you delete the typealias
ItemType = Int line from the code above, everything still works, because it is clear extension Array: Container {}
what type should be used for ItemType. Array’s existing append(_:) method and subscript enable Swift to infer the appropriate
type to use for ItemType, just as for the generic Stack type above. After defining this
You can also make the generic Stack type conform to the Container protocol: extension, you can use any Array as a Container.

struct Stack<Element>: Container {

Generic Where Clauses
// original Stack<Element> implementation
Type constraints, as described in Type Constraints, enable you to define requirements
var items = [Element]() on the type parameters associated with a generic function or type.
mutating func push(_ item: Element) {
items.append(item) It can also be useful to define requirements for associated types. You do this by
}
defining a generic where clause. A generic where clause enables you to require that
an associated type must conform to a certain protocol, or that certain type parameters
mutating func pop() -> Element {
and associated types must be the same. A generic where clause starts with the where
return items.removeLast() keyword, followed by constraints for associated types or equality relationships
} between types and associated types. You write a generic where clause right before the
// conformance to the Container protocol opening curly brace of a type or function’s body.
mutating func append(_ item: Element) {
The example below defines a generic function called allItemsMatch, which checks to
self.push(item)
see if two Container instances contain the same items in the same order. The
} function returns a Boolean value of true if all items match and a value of false if they
var count: Int { do not.
return items.count
}
The two containers to be checked do not have to be the same type of container
(although they can be), but they do have to hold the same type of items. This
subscript(i: Int) -> Element {
requirement is expressed through a combination of type constraints and a generic
return items[i] where clause:
}
func allItemsMatch<C1: Container, C2: Container>
}

169
 (_ someContainer: C1, _ anotherContainer: C2) -> Bool anotherContainer is a container of type C2.
where C1.ItemType == C2.ItemType, C1.ItemType: Equatable {
someContainer and anotherContainer contain the same type of items.

The items in someContainer can be checked with the not equal operator (!
// Check that both containers contain the same number of
=) to see if they are different from each other.
items.
if someContainer.count != anotherContainer.count { The third and fourth requirements combine to mean that the items in
anotherContainer can also be checked with the != operator, because they are exactly
return false
the same type as the items in someContainer.
}

These requirements enable the allItemsMatch(_:_:) function to compare the two

// Check each pair of items to see if they are equivalent. containers, even if they are of a different container type.
for i in 0..<someContainer.count {
if someContainer[i] != anotherContainer[i] { The allItemsMatch(_:_:) function starts by checking that both containers contain the
same number of items. If they contain a different number of items, there is no way that
return false
they can match, and the function returns false.
}
} After making this check, the function iterates over all of the items in someContainer
with a for-in loop and the half-open range operator (..<). For each item, the function
// All items match, so return true. checks whether the item from someContainer is not equal to the corresponding item in
anotherContainer. If the two items are not equal, then the two containers do not
return true
match, and the function returns false.
}

This function takes two arguments called someContainer and anotherContainer. The If the loop finishes without finding a mismatch, the two containers match, and the
someContainer argument is of type C1, and the anotherContainer argument is of type function returns true.
C2. Both C1 and C2 are type parameters for two container types to be determined when
the function is called. Here’s how the allItemsMatch(_:_:) function looks in action:

The following requirements are placed on the function’s two type parameters: var stackOfStrings = Stack<String>()
stackOfStrings.push("uno")
C1 must conform to the Container protocol (written as C1: Container). stackOfStrings.push("dos")
C2 must also conform to the Container protocol (written as C2: stackOfStrings.push("tres")
Container).

The ItemType for C1 must be the same as the ItemType for C2 (written as var arrayOfStrings = ["uno", "dos", "tres"]
C1.ItemType == C2.ItemType).

The ItemType for C1 must conform to the Equatable protocol (written as if allItemsMatch(stackOfStrings, arrayOfStrings) {
C1.ItemType: Equatable). print("All items match.")
The first and second requirements are defined in the function’s type parameter list, } else {
and the third and fourth requirements are defined in the function’s generic where print("Not all items match.")
clause. }
// Prints "All items match."
These requirements mean:
The example above creates a Stack instance to store String values, and pushes
someContainer is a container of type C1. three strings onto the stack. The example also creates an Array instance initialized

170
with an array literal containing the same three strings as the stack. Even though the
stack and the array are of a different type, they both conform to the Container
protocol, and both contain the same type of values. You can therefore call the
allItemsMatch(_:_:) function with these two containers as its arguments. In the
example above, the allItemsMatch(_:_:) function correctly reports that all of the
items in the two containers match.

171
 Section 24 A source file is a single Swift source code file within a module (in effect, a single file
within an app or framework). Although it is common to define individual types in

Access Control separate source files, a single source file can contain definitions for multiple types,
functions, and so on.

Access Levels
Swift provides five different access levels for entities within your code. These access
Access Control levels are relative to the source file in which an entity is defined, and also relative to
the module that source file belongs to.
Access control restricts access to parts of your code from code in other source files
and modules. This feature enables you to hide the implementation details of your Open access and public access enable entities to be used within any
code, and to specify a preferred interface through which that code can be accessed source file from their defining module, and also in a source file from
and used. another module that imports the defining module. You typically use open or
public access when specifying the public interface to a framework. The
You can assign specific access levels to individual types (classes, structures, and difference between open and public access is described below.
enumerations), as well as to properties, methods, initializers, and subscripts
Internal access enables entities to be used within any source file from their
belonging to those types. Protocols can be restricted to a certain context, as can
defining module, but not in any source file outside of that module. You
global constants, variables, and functions.
typically use internal access when defining an app’s or a framework’s
internal structure.
In addition to offering various levels of access control, Swift reduces the need to
specify explicit access control levels by providing default access levels for typical File-private access restricts the use of an entity to its own defining source
scenarios. Indeed, if you are writing a single-target app, you may not need to specify file. Use file-private access to hide the implementation details of a specific
explicit access control levels at all. piece of functionality when those details are used within an entire file.
Private access restricts the use of an entity to the enclosing declaration.
NOTE
Use private access to hide the implementation details of a specific piece of
The various aspects of your code that can have access control applied to them functionality when those details are used only within a single declaration.
(properties, types, functions, and so on) are referred to as “entities” in the sections
below, for brevity. Open access is the highest (least restrictive) access level and private access is the
lowest (most restrictive) access level.
Modules and Source Files
Open access applies only to classes and class members, and it differs from public
Swift’s access control model is based on the concept of modules and source files. access as follows:

A module is a single unit of code distribution—a framework or application that is built Classes with public access, or any more restrictive access level, can be
and shipped as a single unit and that can be imported by another module with Swift’s subclassed only within the module where they’re defined.
import keyword.
Class members with public access, or any more restrictive access level,
Each build target (such as an app bundle or framework) in Xcode is treated as a can be overridden by subclasses only within the module where they’re
separate module in Swift. If you group together aspects of your app’s code as a stand- defined.
alone framework—perhaps to encapsulate and reuse that code across multiple Open classes can be subclassed within the module where they’re defined,
applications—then everything you define within that framework will be part of a and within any module that imports the module where they’re defined.
separate module when it is imported and used within an app, or when it is used within
another framework. Open class members can be overridden by subclasses within the module
where they’re defined, and within any module that imports the module
where they’re defined.

172
Marking a class as open explicitly indicates that you’ve considered the impact of code Access Levels for Unit Test Targets
from other modules using that class as a superclass, and that you’ve designed your When you write an app with a unit test target, the code in your app needs to be made
class’s code accordingly. available to that module in order to be tested. By default, only entities marked as open
or public are accessible to other modules. However, a unit test target can access any
Guiding Principle of Access Levels internal entity, if you mark the import declaration for a product module with the
Access levels in Swift follow an overall guiding principle: No entity can be defined in @testable attribute and compile that product module with testing enabled.
terms of another entity that has a lower (more restrictive) access level.
Access Control Syntax
For example: Define the access level for an entity by placing one of the open, public, internal,
fileprivate, or private modifiers before the entity’s introducer:
A public variable cannot be defined as having an internal, file-private, or
private type, because the type might not be available everywhere that the public class SomePublicClass {}
public variable is used. internal class SomeInternalClass {}
A function cannot have a higher access level than its parameter types and fileprivate class SomeFilePrivateClass {}
return type, because the function could be used in situations where its private class SomePrivateClass {}
constituent types are not available to the surrounding code.
The specific implications of this guiding principle for different aspects of the language public var somePublicVariable = 0
are covered in detail below.
internal let someInternalConstant = 0

Default Access Levels fileprivate func someFilePrivateFunction() {}

All entities in your code (with a few specific exceptions, as described later in this private func somePrivateFunction() {}
chapter) have a default access level of internal if you do not specify an explicit access Unless otherwise specified, the default access level is internal, as described in Default
level yourself. As a result, in many cases you do not need to specify an explicit access Access Levels. This means that SomeInternalClass and someInternalConstant can
level in your code. be written without an explicit access-level modifier, and will still have an access level
of internal:
Access Levels for Single-Target Apps
When you write a simple single-target app, the code in your app is typically self- class SomeInternalClass {} // implicitly internal
contained within the app and does not need to be made available outside of the app’s let someInternalConstant = 0 // implicitly internal
module. The default access level of internal already matches this requirement.
Therefore, you do not need to specify a custom access level. You may, however, want Custom Types
to mark some parts of your code as file private or private in order to hide their If you want to specify an explicit access level for a custom type, do so at the point that
implementation details from other code within the app’s module. you define the type. The new type can then be used wherever its access level permits.
For example, if you define a file-private class, that class can only be used as the type
Access Levels for Frameworks of a property, or as a function parameter or return type, in the source file in which the
When you develop a framework, mark the public-facing interface to that framework as file-private class is defined.
open or public so that it can be viewed and accessed by other modules, such as an
app that imports the framework. This public-facing interface is the application The access control level of a type also affects the default access level of that type’s
programming interface (or API) for the framework. members (its properties, methods, initializers, and subscripts). If you define a type’s
access level as private or file private, the default access level of its members will also
NOTE be private or file private. If you define a type’s access level as internal or public (or use
Any internal implementation details of your framework can still use the default access the default access level of internal without specifying an access level explicitly), the
level of internal, or can be marked as private or file private if you want to hide them default access level of the type’s members will be internal.
from other parts of the framework’s internal code. You need to mark an entity as open
or public only if you want it to become part of your framework’s API.

173
 I M P O R TA N T internal access and one with private access, the access level for that compound tuple
A public type defaults to having internal members, not public members. If you want a type will be private.
type member to be public, you must explicitly mark it as such. This requirement
ensures that the public-facing API for a type is something you opt in to publishing, and NOTE
avoids presenting the internal workings of a type as public API by mistake.
Tuple types do not have a standalone definition in the way that classes, structures,
public class SomePublicClass { // explicitly enumerations, and functions do. A tuple type’s access level is deduced automatically
public class when the tuple type is used, and cannot be specified explicitly.
public var somePublicProperty = 0 // explicitly
public class member Function Types
var someInternalProperty = 0 // implicitly The access level for a function type is calculated as the most restrictive access level
internal class member of the function’s parameter types and return type. You must specify the access level
fileprivate func someFilePrivateMethod() {} // explicitly explicitly as part of the function’s definition if the function’s calculated access level
file-private class member does not match the contextual default.
private func somePrivateMethod() {} // explicitly
private class member The example below defines a global function called someFunction(), without providing
} a specific access-level modifier for the function itself. You might expect this function to
have the default access level of “internal”, but this is not the case. In fact,
someFunction() will not compile as written below:
class SomeInternalClass { // implicitly
internal class
func someFunction() -> (SomeInternalClass, SomePrivateClass) {
var someInternalProperty = 0 // implicitly
internal class member // function implementation goes here
fileprivate func someFilePrivateMethod() {} // explicitly file- }
private class member
The function’s return type is a tuple type composed from two of the custom classes
private func somePrivateMethod() {} // explicitly defined above in Custom Types. One of these classes was defined as “internal”, and
private class member
the other was defined as “private”. Therefore, the overall access level of the
} compound tuple type is “private” (the minimum access level of the tuple’s constituent
types).
fileprivate class SomeFilePrivateClass { // explicitly file-
private class Because the function’s return type is private, you must mark the function’s overall
func someFilePrivateMethod() {} // implicitly file- access level with the private modifier for the function declaration to be valid:
private class member
private func somePrivateMethod() {} // explicitly private func someFunction() -> (SomeInternalClass,
private class member SomePrivateClass) {

} // function implementation goes here

}

private class SomePrivateClass { // explicitly It is not valid to mark the definition of someFunction() with the public or internal
private class modifiers, or to use the default setting of internal, because public or internal users of
func somePrivateMethod() {} // implicitly the function might not have appropriate access to the private class used in the
private class member function’s return type.
}
Enumeration Types
Tuple Types
The individual cases of an enumeration automatically receive the same access level
The access level for a tuple type is the most restrictive access level of all types used as the enumeration they belong to. You cannot specify a different access level for
in that tuple. For example, if you compose a tuple from two different types, one with individual enumeration cases.

174
In the example below, the CompassPoint enumeration has an explicit access level of override internal func someMethod() {}
“public”. The enumeration cases north, south, east, and west therefore also have an }
access level of “public”:
It is even valid for a subclass member to call a superclass member that has lower
access permissions than the subclass member, as long as the call to the superclass’s
public enum CompassPoint {
member takes place within an allowed access level context (that is, within the same
case north source file as the superclass for a file-private member call, or within the same module
case south as the superclass for an internal member call):
case east
public class A {
case west
fileprivate func someMethod() {}
}
}
Raw Values and Associated Values
The types used for any raw values or associated values in an enumeration definition internal class B: A {
must have an access level at least as high as the enumeration’s access level. You override internal func someMethod() {
cannot use a private type as the raw-value type of an enumeration with an internal
super.someMethod()
access level, for example.
}
Nested Types }
Nested types defined within a private type have an automatic access level of private. Because superclass A and subclass B are defined in the same source file, it is valid for
Nested types defined within a file-private type have an automatic access level of file the B implementation of someMethod() to call super.someMethod().
private. Nested types defined within a public type or an internal type have an
automatic access level of internal. If you want a nested type within a public type to be Constants, Variables, Properties, and Subscripts
publicly available, you must explicitly declare the nested type as public.
A constant, variable, or property cannot be more public than its type. It is not valid to
write a public property with a private type, for example. Similarly, a subscript cannot
Subclassing be more public than either its index type or return type.
You can subclass any class that can be accessed in the current access context. A
subclass cannot have a higher access level than its superclass—for example, you If a constant, variable, property, or subscript makes use of a private type, the constant,
cannot write a public subclass of an internal superclass. variable, property, or subscript must also be marked as private:

In addition, you can override any class member (method, property, initializer, or private var privateInstance = SomePrivateClass()
subscript) that is visible in a certain access context.
Getters and Setters
An override can make an inherited class member more accessible than its superclass Getters and setters for constants, variables, properties, and subscripts automatically
version. In the example below, class A is a public class with a file-private method receive the same access level as the constant, variable, property, or subscript they
called someMethod(). Class B is a subclass of A, with a reduced access level of belong to.
“internal”. Nonetheless, class B provides an override of someMethod() with an access
level of “internal”, which is higher than the original implementation of someMethod(): You can give a setter a lower access level than its corresponding getter, to restrict the
read-write scope of that variable, property, or subscript. You assign a lower access
public class A { level by writing fileprivate(set), private(set), or internal(set) before the var or
fileprivate func someMethod() {} subscript introducer.
}
NOTE

internal class B: A {

175
 This rule applies to stored properties as well as computed properties. Even though you Although you can query the current value of the numberOfEdits property from within
do not write an explicit getter and setter for a stored property, Swift still synthesizes an another source file, you cannot modify the property from another source file. This
implicit getter and setter for you to provide access to the stored property’s backing
restriction protects the implementation details of the TrackedString edit-tracking
storage. Use fileprivate(set), private(set), and internal(set) to change the
access level of this synthesized setter in exactly the same way as for an explicit setter functionality, while still providing convenient access to an aspect of that functionality.
in a computed property.
Note that you can assign an explicit access level for both a getter and a setter if
The example below defines a structure called TrackedString, which keeps track of required. The example below shows a version of the TrackedString structure in which
the number of times a string property is modified: the structure is defined with an explicit access level of public. The structure’s members
(including the numberOfEdits property) therefore have an internal access level by
struct TrackedString {
default. You can make the structure’s numberOfEdits property getter public, and its
private(set) var numberOfEdits = 0 property setter private, by combining the public and private(set) access-level
var value: String = "" { modifiers:
didSet {
public struct TrackedString {
numberOfEdits += 1
public private(set) var numberOfEdits = 0
}
public var value: String = "" {
}
didSet {
}
numberOfEdits += 1
The TrackedString structure defines a stored string property called value, with an
}
initial value of "" (an empty string). The structure also defines a stored integer
property called numberOfEdits, which is used to track the number of times that value }
is modified. This modification tracking is implemented with a didSet property observer public init() {}
on the value property, which increments numberOfEdits every time the value property }
is set to a new value.
Initializers
The TrackedString structure and the value property do not provide an explicit
Custom initializers can be assigned an access level less than or equal to the type that
access-level modifier, and so they both receive the default access level of internal.
they initialize. The only exception is for required initializers (as defined in Required
However, the access level for the numberOfEdits property is marked with a
Initializers). A required initializer must have the same access level as the class it
private(set) modifier to indicate that the property’s getter still has the default access
belongs to.
level of internal, but the property is settable only from within code that’s part of the
TrackedString structure. This enables TrackedString to modify the numberOfEdits
property internally, but to present the property as a read-only property when it is used As with function and method parameters, the types of an initializer’s parameters
outside the structure’s definition—including any extensions to TrackedString. cannot be more private than the initializer’s own access level.

If you create a TrackedString instance and modify its string value a few times, you
Default Initializers
can see the numberOfEdits property value update to match the number of As described in Default Initializers, Swift automatically provides a default initializer
modifications: without any arguments for any structure or base class that provides default values for
all of its properties and does not provide at least one initializer itself.
var stringToEdit = TrackedString()
stringToEdit.value = "This string will be tracked." A default initializer has the same access level as the type it initializes, unless that type
is defined as public. For a type that is defined as public, the default initializer is
stringToEdit.value += " This edit will increment numberOfEdits."
considered internal. If you want a public type to be initializable with a no-argument
stringToEdit.value += " So will this one."
initializer when used in another module, you must explicitly provide a public no-
print("The number of edits is \(stringToEdit.numberOfEdits)") argument initializer yourself as part of the type’s definition.
// Prints "The number of edits is 3"

176
Default Memberwise Initializers for Structure Types an internal protocol, the type’s implementation of each protocol requirement must be
The default memberwise initializer for a structure type is considered private if any of at least “internal”.
the structure’s stored properties are private. Likewise, if any of the structure’s stored
properties are file private, the initializer is file private. Otherwise, the initializer has an NOTE

access level of internal. In Swift, as in Objective-C, protocol conformance is global—it is not possible for a type
to conform to a protocol in two different ways within the same program.
As with the default initializer above, if you want a public structure type to be
initializable with a memberwise initializer when used in another module, you must Extensions
provide a public memberwise initializer yourself as part of the type’s definition. You can extend a class, structure, or enumeration in any access context in which the
class, structure, or enumeration is available. Any type members added in an extension
Protocols have the same default access level as type members declared in the original type
If you want to assign an explicit access level to a protocol type, do so at the point that being extended. If you extend a public or internal type, any new type members you
you define the protocol. This enables you to create protocols that can only be adopted add have a default access level of internal. If you extend a file-private type, any new
within a certain access context. type members you add have a default access level of file private. If you extend a
private type, any new type members you add have a default access level of private.
The access level of each requirement within a protocol definition is automatically set
to the same access level as the protocol. You cannot set a protocol requirement to a Alternatively, you can mark an extension with an explicit access-level modifier (for
different access level than the protocol it supports. This ensures that all of the example, private extension) to set a new default access level for all members
protocol’s requirements will be visible on any type that adopts the protocol. defined within the extension. This new default can still be overridden within the
extension for individual type members.
NOTE

If you define a public protocol, the protocol’s requirements require a public access level
Adding Protocol Conformance with an Extension
for those requirements when they are implemented. This behavior is different from You cannot provide an explicit access-level modifier for an extension if you are using
other types, where a public type definition implies an access level of internal for the that extension to add protocol conformance. Instead, the protocol’s own access level
type’s members. is used to provide the default access level for each protocol requirement
implementation within the extension.
Protocol Inheritance
If you define a new protocol that inherits from an existing protocol, the new protocol Generics
can have at most the same access level as the protocol it inherits from. You cannot
write a public protocol that inherits from an internal protocol, for example. The access level for a generic type or generic function is the minimum of the access
level of the generic type or function itself and the access level of any type constraints
Protocol Conformance on its type parameters.

A type can conform to a protocol with a lower access level than the type itself. For
example, you can define a public type that can be used in other modules, but whose Type Aliases
conformance to an internal protocol can only be used within the internal protocol’s Any type aliases you define are treated as distinct types for the purposes of access
defining module. control. A type alias can have an access level less than or equal to the access level of
the type it aliases. For example, a private type alias can alias a private, file-private,
The context in which a type conforms to a particular protocol is the minimum of the internal, public, or open type, but a public type alias cannot alias an internal, file-
type’s access level and the protocol’s access level. If a type is public, but a protocol it private, or private type.
conforms to is internal, the type’s conformance to that protocol is also internal.
NOTE

When you write or extend a type to conform to a protocol, you must ensure that the This rule also applies to type aliases for associated types used to satisfy protocol
type’s implementation of each protocol requirement has at least the same access level conformances.
as the type’s conformance to that protocol. For example, if a public type conforms to

177
 Section 25

Advanced Operators
Advanced Operators The bitwise NOT operator is a prefix operator, and appears immediately before the
value it operates on, without any white space:
In addition to the operators described in Basic Operators, Swift provides several
advanced operators that perform more complex value manipulation. These include all let initialBits: UInt8 = 0b00001111
of the bitwise and bit shifting operators you will be familiar with from C and Objective- let invertedBits = ~initialBits // equals 11110000
C. UInt8 integers have eight bits and can store any value between 0 and 255. This
example initializes a UInt8 integer with the binary value 00001111, which has its first
Unlike arithmetic operators in C, arithmetic operators in Swift do not overflow by four bits set to 0, and its second four bits set to 1. This is equivalent to a decimal value
default. Overflow behavior is trapped and reported as an error. To opt in to overflow of 15.
behavior, use Swift’s second set of arithmetic operators that overflow by default, such
as the overflow addition operator (&+). All of these overflow operators begin with an The bitwise NOT operator is then used to create a new constant called invertedBits,
ampersand (&). which is equal to initialBits, but with all of the bits inverted. Zeros become ones,
and ones become zeros. The value of invertedBits is 11110000, which is equal to an
When you define your own structures, classes, and enumerations, it can be useful to unsigned decimal value of 240.
provide your own implementations of the standard Swift operators for these custom
types. Swift makes it easy to provide tailored implementations of these operators and Bitwise AND Operator
to determine exactly what their behavior should be for each type you create.
The bitwise AND operator (&) combines the bits of two numbers. It returns a new
number whose bits are set to 1 only if the bits were equal to 1 in both input numbers:
You’re not limited to the predefined operators. Swift gives you the freedom to define
your own custom infix, prefix, postfix, and assignment operators, with custom
precedence and associativity values. These operators can be used and adopted in
your code like any of the predefined operators, and you can even extend existing
types to support the custom operators you define.

Bitwise Operators
Bitwise operators enable you to manipulate the individual raw data bits within a data
structure. They are often used in low-level programming, such as graphics
programming and device driver creation. Bitwise operators can also be useful when
you work with raw data from external sources, such as encoding and decoding data
for communication over a custom protocol.
In the example below, the values of firstSixBits and lastSixBits both have four
Swift supports all of the bitwise operators found in C, as described below. middle bits equal to 1. The bitwise AND operator combines them to make the number
00111100, which is equal to an unsigned decimal value of 60:
Bitwise NOT Operator
The bitwise NOT operator (~) inverts all bits in a number: let firstSixBits: UInt8 = 0b11111100
let lastSixBits: UInt8 = 0b00111111
let middleFourBits = firstSixBits & lastSixBits // equals 00111100

178
Bitwise OR Operator let otherBits: UInt8 = 0b00000101

The bitwise OR operator (|) compares the bits of two numbers. The operator returns a let outputBits = firstBits ^ otherBits // equals 00010001
new number whose bits are set to 1 if the bits are equal to 1 in either input number:
Bitwise Left and Right Shift Operators
The bitwise left shift operator (<<) and bitwise right shift operator (>>) move all bits in a
number to the left or the right by a certain number of places, according to the rules
defined below.

Bitwise left and right shifts have the effect of multiplying or dividing an integer by a
factor of two. Shifting an integer’s bits to the left by one position doubles its value,
whereas shifting it to the right by one position halves its value.

Shifting Behavior for Unsigned Integers

The bit-shifting behavior for unsigned integers is as follows:

In the example below, the values of someBits and moreBits have different bits set to 1.Existing bits are moved to the left or right by the requested number of
1. The bitwise OR operator combines them to make the number 11111110, which places.
equals an unsigned decimal of 254:
2.Any bits that are moved beyond the bounds of the integer’s storage are
let someBits: UInt8 = 0b10110010
discarded.

let moreBits: UInt8 = 0b01011110 3.Zeros are inserted in the spaces left behind after the original bits are moved
let combinedbits = someBits | moreBits // equals 11111110
to the left or right.
This approach is known as a logical shift.
Bitwise XOR Operator
The bitwise XOR operator, or “exclusive OR operator” (^), compares the bits of two The illustration below shows the results of 11111111 << 1 (which is 11111111 shifted
numbers. The operator returns a new number whose bits are set to 1 where the input to the left by 1 place), and 11111111 >> 1 (which is 11111111 shifted to the right by 1
bits are different and are set to 0 where the input bits are the same: place). Blue numbers are shifted, gray numbers are discarded, and orange zeros are
inserted:

Here’s how bit shifting looks in Swift code:

let shiftBits: UInt8 = 4 // 00000100 in binary

shiftBits << 1 // 00001000
In the example below, the values of firstBits and otherBits each have a bit set to 1
shiftBits << 2 // 00010000
in a location that the other does not. The bitwise XOR operator sets both of these bits
to 1 in its output value. All of the other bits in firstBits and otherBits match and are shiftBits << 5 // 10000000
set to 0 in the output value: shiftBits << 6 // 00000000
shiftBits >> 2 // 00000001
let firstBits: UInt8 = 0b00010100

179
You can use bit shifting to encode and decode values within other data types: The remaining bits (known as the value bits) store the actual value. Positive numbers
are stored in exactly the same way as for unsigned integers, counting upwards from 0.
let pink: UInt32 = 0xCC6699 Here’s how the bits inside an Int8 look for the number 4:
let redComponent = (pink & 0xFF0000) >> 16 // redComponent is
0xCC, or 204
let greenComponent = (pink & 0x00FF00) >> 8 // greenComponent is
0x66, or 102
let blueComponent = pink & 0x0000FF // blueComponent is
0x99, or 153

This example uses a UInt32 constant called pink to store a Cascading Style Sheets
color value for the color pink. The CSS color value #CC6699 is written as 0xCC6699 in
Swift’s hexadecimal number representation. This color is then decomposed into its red The sign bit is 0 (meaning “positive”), and the seven value bits are just the number 4,
(CC), green (66), and blue (99) components by the bitwise AND operator (&) and the written in binary notation.
bitwise right shift operator (>>).
Negative numbers, however, are stored differently. They are stored by subtracting
The red component is obtained by performing a bitwise AND between the numbers their absolute value from 2 to the power of n, where n is the number of value bits. An
0xCC6699 and 0xFF0000. The zeros in 0xFF0000 effectively “mask” the second and third eight-bit number has seven value bits, so this means 2 to the power of 7, or 128.
bytes of 0xCC6699, causing the 6699 to be ignored and leaving 0xCC0000 as the result.
Here’s how the bits inside an Int8 look for the number -4:
This number is then shifted 16 places to the right (>> 16). Each pair of characters in a
hexadecimal number uses 8 bits, so a move 16 places to the right will convert
0xCC0000 into 0x0000CC. This is the same as 0xCC, which has a decimal value of 204.

Similarly, the green component is obtained by performing a bitwise AND between the
numbers 0xCC6699 and 0x00FF00, which gives an output value of 0x006600. This
output value is then shifted eight places to the right, giving a value of 0x66, which has
a decimal value of 102.
This time, the sign bit is 1 (meaning “negative”), and the seven value bits have a
Finally, the blue component is obtained by performing a bitwise AND between the binary value of 124 (which is 128 - 4):
numbers 0xCC6699 and 0x0000FF, which gives an output value of 0x000099. There’s no
need to shift this to the right, as 0x000099 already equals 0x99, which has a decimal
value of 153.

Shifting Behavior for Signed Integers

The shifting behavior is more complex for signed integers than for unsigned integers,
because of the way signed integers are represented in binary. (The examples below
are based on 8-bit signed integers for simplicity, but the same principles apply for
signed integers of any size.) This encoding for negative numbers is known as a two’s complement representation.
It may seem an unusual way to represent negative numbers, but it has several
advantages.
Signed integers use their first bit (known as the sign bit) to indicate whether the
integer is positive or negative. A sign bit of 0 means positive, and a sign bit of 1 means
negative. First, you can add -1 to -4, simply by performing a standard binary addition of all eight
bits (including the sign bit), and discarding anything that doesn’t fit in the eight bits
once you’re done:

180
 potentialOverflow += 1
// this causes an error

Providing error handling when values get too large or too small gives you much more
flexibility when coding for boundary value conditions.

However, when you specifically want an overflow condition to truncate the number of
available bits, you can opt in to this behavior rather than triggering an error. Swift
provides three arithmetic overflow operators that opt in to the overflow behavior for
integer calculations. These operators all begin with an ampersand (&):

Second, the two’s complement representation also lets you shift the bits of negative
Overflow addition (&+)
numbers to the left and right like positive numbers, and still end up doubling them for
every shift you make to the left, or halving them for every shift you make to the right. Overflow subtraction (&-)
To achieve this, an extra rule is used when signed integers are shifted to the right: Overflow multiplication (&*)
When you shift signed integers to the right, apply the same rules as for unsigned
integers, but fill any empty bits on the left with the sign bit, rather than with a zero. Value Overflow
Numbers can overflow in both the positive and negative direction.

Here’s an example of what happens when an unsigned integer is allowed to overflow

in the positive direction, using the overflow addition operator (&+):

var unsignedOverflow = UInt8.max

// unsignedOverflow equals 255, which is the maximum value a UInt8
This action ensures that signed integers have the same sign after they are shifted to can hold
the right, and is known as an arithmetic shift. unsignedOverflow = unsignedOverflow &+ 1
// unsignedOverflow is now equal to 0
Because of the special way that positive and negative numbers are stored, shifting
either of them to the right moves them closer to zero. Keeping the sign bit the same The variable unsignedOverflow is initialized with the maximum value a UInt8 can hold
during this shift means that negative integers remain negative as their value moves (255, or 11111111 in binary). It is then incremented by 1 using the overflow addition
closer to zero. operator (&+). This pushes its binary representation just over the size that a UInt8 can
hold, causing it to overflow beyond its bounds, as shown in the diagram below. The
value that remains within the bounds of the UInt8 after the overflow addition is
Overflow Operators 00000000, or zero.
If you try to insert a number into an integer constant or variable that cannot hold that
value, by default Swift reports an error rather than allowing an invalid value to be
created. This behavior gives extra safety when you work with numbers that are too
large or too small.

For example, the Int16 integer type can hold any signed integer between -32768 and
32767. Trying to set an Int16 constant or variable to a number outside of this range
causes an error:

var potentialOverflow = Int16.max

// potentialOverflow equals 32767, which is the maximum value an Something similar happens when an unsigned integer is allowed to overflow in the
Int16 can hold negative direction. Here’s an example using the overflow subtraction operator (&-):

181
 var unsignedOverflow = UInt8.min For both signed and unsigned integers, overflow in the positive direction wraps around
// unsignedOverflow equals 0, which is the minimum value a UInt8 from the maximum valid integer value back to the minimum, and overflow in the
can hold negative direction wraps around from the minimum value to the maximum.
unsignedOverflow = unsignedOverflow &- 1
// unsignedOverflow is now equal to 255 Precedence and Associativity
The minimum value that a UInt8 can hold is zero, or 00000000 in binary. If you Operator precedence gives some operators higher priority than others; these
subtract 1 from 00000000 using the overflow subtraction operator (&-), the number will operators are applied first.
overflow and wrap around to 11111111, or 255 in decimal.
Operator associativity defines how operators of the same precedence are grouped
together—either grouped from the left, or grouped from the right. Think of it as
meaning “they associate with the expression to their left,” or “they associate with the
expression to their right.”

It is important to consider each operator’s precedence and associativity when working

out the order in which a compound expression will be calculated. For example,
operator precedence explains why the following expression equals 17.

2 + 3 % 4 * 5
Overflow also occurs for signed integers. All addition and subtraction for signed // this equals 17
integers is performed in bitwise fashion, with the sign bit included as part of the
If you read strictly from left to right, you might expect the expression to be calculated
numbers being added or subtracted, as described in Bitwise Left and Right Shift
as follows:
Operators.

var signedOverflow = Int8.min 2 plus 3 equals 5

// signedOverflow equals -128, which is the minimum value an Int8 5 remainder 4 equals 1
can hold
1 times 5 equals 5
signedOverflow = signedOverflow &- 1
However, the actual answer is 17, not 5. Higher-precedence operators are evaluated
// signedOverflow is now equal to 127
before lower-precedence ones. In Swift, as in C, the remainder operator (%) and the
The minimum value that an Int8 can hold is -128, or 10000000 in binary. Subtracting 1 multiplication operator (*) have a higher precedence than the addition operator (+). As
from this binary number with the overflow operator gives a binary value of 01111111, a result, they are both evaluated before the addition is considered.
which toggles the sign bit and gives positive 127, the maximum positive value that an
Int8 can hold. However, remainder and multiplication have the same precedence as each other. To
work out the exact evaluation order to use, you also need to consider their
associativity. Remainder and multiplication both associate with the expression to their
left. Think of this as adding implicit parentheses around these parts of the expression,
starting from their left:

2 + ((3 % 4) * 5)

(3 % 4) is 3, so this is equivalent to:

2 + (3 * 5)

(3 * 5) is 15, so this is equivalent to:

2 + 15

182
This calculation yields the final answer of 17. method returns a new Vector2D instance, whose x and y properties are initialized with
the sum of the x and y properties from the two Vector2D instances that are added
For a complete list of Swift operator precedences and associativity rules, see together.
Expressions. For information about the operators provided by the Swift standard
library, see Swift Standard Library Operators Reference. The type method can be used as an infix operator between existing Vector2D
instances:
NOTE
let vector = Vector2D(x: 3.0, y: 1.0)
Swift’s operator precedences and associativity rules are simpler and more predictable
than those found in C and Objective-C. However, this means that they are not exactly let anotherVector = Vector2D(x: 2.0, y: 4.0)
the same as in C-based languages. Be careful to ensure that operator interactions still let combinedVector = vector + anotherVector
behave in the way you intend when porting existing code to Swift.
// combinedVector is a Vector2D instance with values of (5.0, 5.0)

Operator Methods This example adds together the vectors (3.0, 1.0) and (2.0, 4.0) to make the
Classes and structures can provide their own implementations of existing operators. vector (5.0, 5.0), as illustrated below.
This is known as overloading the existing operators.

The example below shows how to implement the arithmetic addition operator (+) for a
custom structure. The arithmetic addition operator is a binary operator because it
operates on two targets and is said to be infix because it appears in between those
two targets.

The example defines a Vector2D structure for a two-dimensional position vector (x,
y), followed by a definition of an operator method to add together instances of the
Vector2D structure:

struct Vector2D {
var x = 0.0, y = 0.0
}

extension Vector2D {
static func + (left: Vector2D, right: Vector2D) -> Vector2D {
return Vector2D(x: left.x + right.x, y: left.y + right.y)
}
}
Prefix and Postfix Operators
The operator method is defined as a type method on Vector2D, with a method name
that matches the operator to be overloaded (+). Because addition isn’t part of the The example shown above demonstrates a custom implementation of a binary infix
essential behavior for a vector, the type method is defined in an extension of Vector2D operator. Classes and structures can also provide implementations of the standard
rather than in the main structure declaration of Vector2D. Because the arithmetic unary operators. Unary operators operate on a single target. They are prefix if they
addition operator is a binary operator, this operator method takes two input precede their target (such as -a) and postfix operators if they follow their target (such
parameters of type Vector2D and returns a single output value, also of type Vector2D. as b!).

In this implementation, the input parameters are named left and right to represent You implement a prefix or postfix unary operator by writing the prefix or postfix
the Vector2D instances that will be on the left side and right side of the + operator. The modifier before the func keyword when declaring the operator method:

183
 extension Vector2D { // original now has values of (4.0, 6.0)
static prefix func - (vector: Vector2D) -> Vector2D { NOTE
return Vector2D(x: -vector.x, y: -vector.y) It is not possible to overload the default assignment operator (=). Only the compound
} assignment operators can be overloaded. Similarly, the ternary conditional operator
(a ? b : c) cannot be overloaded.
}

The example above implements the unary minus operator (-a) for Vector2D instances. Equivalence Operators
The unary minus operator is a prefix operator, and so this method has to be qualified Custom classes and structures do not receive a default implementation of the
with the prefix modifier. equivalence operators, known as the “equal to” operator (==) and “not equal to”
operator (!=). It is not possible for Swift to guess what would qualify as “equal” for your
For simple numeric values, the unary minus operator converts positive numbers into own custom types, because the meaning of “equal” depends on the roles that those
their negative equivalent and vice versa. The corresponding implementation for types play in your code.
Vector2D instances performs this operation on both the x and y properties:
To use the equivalence operators to check for equivalence of your own custom type,
let positive = Vector2D(x: 3.0, y: 4.0) provide an implementation of the operators in the same way as for other infix
let negative = -positive operators:
// negative is a Vector2D instance with values of (-3.0, -4.0)
extension Vector2D {
let alsoPositive = -negative
static func == (left: Vector2D, right: Vector2D) -> Bool {
// alsoPositive is a Vector2D instance with values of (3.0, 4.0)
return (left.x == right.x) && (left.y == right.y)
Compound Assignment Operators }
Compound assignment operators combine assignment (=) with another operation. For static func != (left: Vector2D, right: Vector2D) -> Bool {
example, the addition assignment operator (+=) combines addition and assignment return !(left == right)
into a single operation. You mark a compound assignment operator’s left input
}
parameter type as inout, because the parameter’s value will be modified directly from
within the operator method. }

The above example implements an “equal to” operator (==) to check if two Vector2D
The example below implements an addition assignment operator method for Vector2D instances have equivalent values. In the context of Vector2D, it makes sense to
instances: consider “equal” as meaning “both instances have the same x values and y values”,
and so this is the logic used by the operator implementation. The example also
extension Vector2D { implements the “not equal to” operator (!=), which simply returns the inverse of the
static func += (left: inout Vector2D, right: Vector2D) { result of the “equal to” operator.
left = left + right
}
You can now use these operators to check whether two Vector2D instances are
equivalent:
}

Because an addition operator was defined earlier, you don’t need to reimplement the let twoThree = Vector2D(x: 2.0, y: 3.0)
addition process here. Instead, the addition assignment operator method takes let anotherTwoThree = Vector2D(x: 2.0, y: 3.0)
advantage of the existing addition operator method, and uses it to set the left value to if twoThree == anotherTwoThree {
be the left value plus the right value:
print("These two vectors are equivalent.")
var original = Vector2D(x: 1.0, y: 2.0) }
let vectorToAdd = Vector2D(x: 3.0, y: 4.0) // Prints "These two vectors are equivalent."
original += vectorToAdd

184
 extension Vector2D {
Custom Operators
static func +- (left: Vector2D, right: Vector2D) -> Vector2D {
You can declare and implement your own custom operators in addition to the standard
operators provided by Swift. For a list of characters that can be used to define custom return Vector2D(x: left.x + right.x, y: left.y - right.y)
operators, see Operators. }
}
New operators are declared at a global level using the operator keyword, and are let firstVector = Vector2D(x: 1.0, y: 2.0)
marked with the prefix, infix or postfix modifiers:
let secondVector = Vector2D(x: 3.0, y: 4.0)
prefix operator +++ let plusMinusVector = firstVector +- secondVector

The example above defines a new prefix operator called +++. This operator does not // plusMinusVector is a Vector2D instance with values of (4.0, -2.0)
have an existing meaning in Swift, and so it is given its own custom meaning below in This operator adds together the x values of two vectors, and subtracts the y value of
the specific context of working with Vector2D instances. For the purposes of this the second vector from the first. Because it is in essence an “additive” operator, it has
example, +++ is treated as a new “prefix doubling” operator. It doubles the x and y been given the same precedence group as additive infix operators such as + and -.
values of a Vector2D instance, by adding the vector to itself with the addition For a complete list of the operator precedence groups and associativity settings, for
assignment operator defined earlier. To implement the +++ operator, you add a type the operators provided by the Swift standard library, see Swift Standard Library
method called +++ to Vector2D as follows: Operators Reference. For more information about precedence groups and to see the
syntax for defining your own operators and precedence groups, see Operator
extension Vector2D { Declaration.
static prefix func +++ (vector: inout Vector2D) -> Vector2D {
vector += vector NOTE

return vector You do not specify a precedence when defining a prefix or postfix operator. However, if
you apply both a prefix and a postfix operator to the same operand, the postfix operator
}
is applied first.
}

var toBeDoubled = Vector2D(x: 1.0, y: 4.0)

let afterDoubling = +++toBeDoubled
// toBeDoubled now has values of (2.0, 8.0)
// afterDoubling also has values of (2.0, 8.0)

Precedence for Custom Infix Operators

Custom infix operators each belong to a precedence group. A precedence group
specifies an operator’s precedence relative to other infix operators, as well as the
operator’s associativity. See Precedence and Associativity for an explanation of how
these characteristics affect an infix operator’s interaction with other infix operators.

A custom infix operator that is not explicitly placed into a precedence group is given a
default precedence group with a precedence immediately higher than the precedence
of the ternary conditional operator.

The following example defines a new custom infix operator called +-, which belongs to
the precedence group AdditionPrecedence:

infix operator +-: AdditionPrecedence

185
 Language Reference
Language Reference

186
 Section 1 This definition indicates that a getter-setter block can consist of a getter clause
followed by an optional setter clause, enclosed in braces, or a setter clause followed

About the Language by a getter clause, enclosed in braces. The grammar production above is equivalent
to the following two productions, where the alternatives are spelled out explicitly:

GRAMMAR OF A GETTER-SETTER BLOCK

getter-setter-block → {getter-clausesetter-clauseopt}
About the Language Reference getter-setter-block → {setter-clausegetter-clause}

This part of the book describes the formal grammar of the Swift programming
language. The grammar described here is intended to help you understand the
language in more detail, rather than to allow you to directly implement a parser or
compiler.

The Swift language is relatively small, because many common types, functions, and
operators that appear virtually everywhere in Swift code are actually defined in the
Swift standard library. Although these types, functions, and operators are not part of
the Swift language itself, they are used extensively in the discussions and code
examples in this part of the book.

How to Read the Grammar

The notation used to describe the formal grammar of the Swift programming language
follows a few conventions:

An arrow (→) is used to mark grammar productions and can be read as

“can consist of.”
Syntactic categories are indicated by italic text and appear on both sides of
a grammar production rule.
Literal words and punctuation are indicated by boldface constant width
text and appear only on the right-hand side of a grammar production rule.
Alternative grammar productions are separated by vertical bars (|). When
alternative productions are too long to read easily, they are broken into
multiple grammar production rules on new lines.
In a few cases, regular font text is used to describe the right-hand side of a
grammar production rule.
Optional syntactic categories and literals are marked by a trailing subscript,
opt.
As an example, the grammar of a getter-setter block is defined as follows:

GRAMMAR OF A GETTER-SETTER BLOCK

getter-setter-block → {getter-clausesetter-clauseopt} {setter-clausegetter-clause}

187
 Section 2 Inside a closure with no explicit parameter names, the parameters are implicitly
named $0, $1, $2, and so on. These names are valid identifiers within the scope of the

Lexical Structure closure.

GRAMMAR OF AN IDENTIFIER

identifier → identifier-headidentifier-charactersopt
identifier → `identifier-headidentifier-charactersopt`
Lexical Structure identifier → implicit-parameter-name
identifier-list → identifier identifier,identifier-list
The lexical structure of Swift describes what sequence of characters form valid tokens identifier-head → Upper- or lowercase letter A through Z
of the language. These valid tokens form the lowest-level building blocks of the identifier-head → _
language and are used to describe the rest of the language in subsequent chapters. A identifier-head → U+00A8, U+00AA, U+00AD, U+00AF, U+00B2–U+00B5, or
token consists of an identifier, keyword, punctuation, literal, or operator. U+00B7–U+00BA
identifier-head → U+00BC–U+00BE, U+00C0–U+00D6, U+00D8–U+00F6, or
In most cases, tokens are generated from the characters of a Swift source file by U+00F8–U+00FF
considering the longest possible substring from the input text, within the constraints of identifier-head → U+0100–U+02FF, U+0370–U+167F, U+1681–U+180D, or
the grammar that are specified below. This behavior is referred to as longest match or U+180F–U+1DBF
maximal munch. identifier-head → U+1E00–U+1FFF
identifier-head → U+200B–U+200D, U+202A–U+202E, U+203F–U+2040, U+2054,
Whitespace and Comments or U+2060–U+206F
identifier-head → U+2070–U+20CF, U+2100–U+218F, U+2460–U+24FF, or
Whitespace has two uses: to separate tokens in the source file and to help determine
U+2776–U+2793
whether an operator is a prefix or postfix (see Operators), but is otherwise ignored.
identifier-head → U+2C00–U+2DFF or U+2E80–U+2FFF
The following characters are considered whitespace: space (U+0020), line feed
identifier-head → U+3004–U+3007, U+3021–U+302F, U+3031–U+303F, or
(U+000A), carriage return (U+000D), horizontal tab (U+0009), vertical tab (U+000B),
U+3040–U+D7FF
form feed (U+000C) and null (U+0000).
identifier-head → U+F900–U+FD3D, U+FD40–U+FDCF, U+FDF0–U+FE1F, or
U+FE30–U+FE44
Comments are treated as whitespace by the compiler. Single line comments begin
identifier-head → U+FE47–U+FFFD
with // and continue until a line feed (U+000A) or carriage return (U+000D). Multiline
identifier-head → U+10000–U+1FFFD, U+20000–U+2FFFD, U+30000–U+3FFFD,
comments begin with /* and end with */. Nesting multiline comments is allowed, but
or U+40000–U+4FFFD
the comment markers must be balanced.
identifier-head → U+50000–U+5FFFD, U+60000–U+6FFFD, U+70000–U+7FFFD,
or U+80000–U+8FFFD
Comments can contain additional formatting and markup, as described in Markup identifier-head → U+90000–U+9FFFD, U+A0000–U+AFFFD, U+B0000–
Formatting Reference. U+BFFFD, or U+C0000–U+CFFFD
identifier-head → U+D0000–U+DFFFD or U+E0000–U+EFFFD
Identifiers identifier-character → Digit 0 through 9
Identifiers begin with an uppercase or lowercase letter A through Z, an underscore (_), identifier-character → U+0300–U+036F, U+1DC0–U+1DFF, U+20D0–U+20FF, or
a noncombining alphanumeric Unicode character in the Basic Multilingual Plane, or a U+FE20–U+FE2F
character outside the Basic Multilingual Plane that isn’t in a Private Use Area. After identifier-character → identifier-head
the first character, digits and combining Unicode characters are also allowed. identifier-characters → identifier-characteridentifier-charactersopt
implicit-parameter-name → $decimal-digits
To use a reserved word as an identifier, put a backtick (`) before and after it. For
example, class is not a valid identifier, but `class` is valid. The backticks are not Keywords and Punctuation
considered part of the identifier; `x` and x have the same meaning. The following keywords are reserved and can’t be used as identifiers, unless they’re
escaped with backticks, as described above in Identifiers. Keywords other than inout,

188
var, and let can be used as parameter names in a function declaration or function A literal doesn’t have a type on its own. Instead, a literal is parsed as having infinite
call without being escaped with backticks. When a member has the same name as a precision and Swift’s type inference attempts to infer a type for the literal. For
keyword, references to that member don’t need to be escaped with backticks, except example, in the declaration let x: Int8 = 42, Swift uses the explicit type annotation
when there is ambiguity between referring to the member and using the keyword—for (: Int8) to infer that the type of the integer literal 42 is Int8. If there isn’t suitable type
example, self, Type, and Protocol have special meaning in an explicit member information available, Swift infers that the literal’s type is one of the default literal types
expression, so they must be escaped with backticks in that context. defined in the Swift standard library. The default types are Int for integer literals,
Double for floating-point literals, String for string literals, and Bool for Boolean literals.
Keywords used in declarations: associatedtype, class, deinit, enum, For example, in the declaration let str = "Hello, world", the default inferred type
extension, fileprivate, func, import, init, inout, internal, let, open, of the string literal "Hello, world" is String.
operator, private, protocol, public, static, struct, subscript,
typealias, and var. When specifying the type annotation for a literal value, the annotation’s type must be a
type that can be instantiated from that literal value. That is, the type must conform to
Keywords used in statements: break, case, continue, default, defer, do,
one of the following Swift standard library protocols: ExpressibleByIntegerLiteral
else, fallthrough, for, guard, if, in, repeat, return, switch, where, and
for integer literals, ExpressibleByFloatLiteral for floating-point literals,
while.
ExpressibleByStringLiteral for string literals, ExpressibleByBooleanLiteral for
Keywords used in expressions and types: as, Any, catch, false, is, nil, Boolean literals, ExpressibleByUnicodeScalarLiteral for string literals that contain
rethrows, super, self, Self, throw, throws, true, and try. only a single Unicode scalar, and ExpressibleByExtendedGraphemeClusterLiteral
Keywords used in patterns: _. for string literals that contain only a single extended grapheme cluster. For example,
Int8 conforms to the ExpressibleByIntegerLiteral protocol, and therefore it can be
Keywords that begin with a number sign (#): #available, #colorLiteral, used in the type annotation for the integer literal 42 in the declaration let x: Int8 =
#column, #else, #elseif, #endif, #file, #fileLiteral, #function, #if, 42.
#imageLiteral, #line, #selector. and #sourceLocation.

Keywords reserved in particular contexts: associativity, convenience, GRAMMAR OF A LITERAL

dynamic, didSet, final, get, infix, indirect, lazy, left, mutating, none, literal → numeric-literal string-literal boolean-literal nil-literal
nonmutating, optional, override, postfix, precedence, prefix, Protocol, numeric-literal → -optinteger-literal -optfloating-point-literal
required, right, set, Type, unowned, weak, and willSet. Outside the boolean-literal → true false
context in which they appear in the grammar, they can be used as nil-literal → nil
identifiers.
The following tokens are reserved as punctuation and can’t be used as custom Integer Literals
operators: (, ), {, }, [, ], ., ,, :, ;, =, @, #, & (as a prefix operator), ->, `, ?, and ! (as a Integer literals represent integer values of unspecified precision. By default, integer
postfix operator). literals are expressed in decimal; you can specify an alternate base using a prefix.
Binary literals begin with 0b, octal literals begin with 0o, and hexadecimal literals begin
Literals with 0x.
A literal is the source code representation of a value of a type, such as a number or
Decimal literals contain the digits 0 through 9. Binary literals contain 0 and 1, octal
string.
literals contain 0 through 7, and hexadecimal literals contain 0 through 9 as well as A
through F in upper- or lowercase.
The following are examples of literals:

42 // Integer literal
Negative integers literals are expressed by prepending a minus sign (-) to an integer
literal, as in -42.
3.14159 // Floating-point literal
"Hello, world!" // String literal Underscores (_) are allowed between digits for readability, but they are ignored and
true // Boolean literal therefore don’t affect the value of the literal. Integer literals can begin with leading
zeros (0), but they are likewise ignored and don’t affect the base or value of the literal.

189
Unless otherwise specified, the default inferred type of an integer literal is the Swift exponent consists of an upper- or lowercase p prefix followed by a sequence of
standard library type Int. The Swift standard library also defines types for various decimal digits that indicates what power of 2 the value preceding the p is multiplied by.
sizes of signed and unsigned integers, as described in Integers. For example, 0xFp2 represents 15 x 22, which evaluates to 60. Similarly, 0xFp-2
represents 15 x 2-2, which evaluates to 3.75.
GRAMMAR OF AN INTEGER LITERAL

integer-literal → binary-literal Negative floating-point literals are expressed by prepending a minus sign (-) to a
integer-literal → octal-literal floating-point literal, as in -42.5.
integer-literal → decimal-literal
integer-literal → hexadecimal-literal Underscores (_) are allowed between digits for readability, but are ignored and
binary-literal → 0bbinary-digitbinary-literal-charactersopt therefore don’t affect the value of the literal. Floating-point literals can begin with
binary-digit → Digit 0 or 1 leading zeros (0), but are likewise ignored and don’t affect the base or value of the
binary-literal-character → binary-digit _ literal.
binary-literal-characters → binary-literal-characterbinary-literal-charactersopt
octal-literal → 0ooctal-digitoctal-literal-charactersopt Unless otherwise specified, the default inferred type of a floating-point literal is the
octal-digit → Digit 0 through 7 Swift standard library type Double, which represents a 64-bit floating-point number.
octal-literal-character → octal-digit _ The Swift standard library also defines a Float type, which represents a 32-bit
octal-literal-characters → octal-literal-characteroctal-literal-charactersopt floating-point number.
decimal-literal → decimal-digitdecimal-literal-charactersopt
decimal-digit → Digit 0 through 9 G R A M M A R O F A F L O AT I N G - P O I N T L I T E R A L

decimal-digits → decimal-digitdecimal-digitsopt floating-point-literal → decimal-literaldecimal-fractionoptdecimal-exponentopt

decimal-literal-character → decimal-digit _ floating-point-literal → hexadecimal-literalhexadecimal-fractionopthexadecimal-
decimal-literal-characters → decimal-literal-characterdecimal-literal-charactersopt exponent
hexadecimal-literal → 0xhexadecimal-digithexadecimal-literal-charactersopt decimal-fraction → .decimal-literal
hexadecimal-digit → Digit 0 through 9, a through f, or A through F decimal-exponent → floating-point-esignoptdecimal-literal
hexadecimal-literal-character → hexadecimal-digit _ hexadecimal-fraction → .hexadecimal-digithexadecimal-literal-charactersopt
hexadecimal-literal-characters → hexadecimal-literal-characterhexadecimal-literal- hexadecimal-exponent → floating-point-psignoptdecimal-literal
charactersopt floating-point-e → e E
floating-point-p → p P
Floating-Point Literals sign → + -
Floating-point literals represent floating-point values of unspecified precision.
String Literals
By default, floating-point literals are expressed in decimal (with no prefix), but they can A string literal is a sequence of characters surrounded by double quotes, with the
also be expressed in hexadecimal (with a 0x prefix). following form:

Decimal floating-point literals consist of a sequence of decimal digits followed by "characters"

either a decimal fraction, a decimal exponent, or both. The decimal fraction consists of String literals cannot contain an unescaped double quote ("), an unescaped backslash
a decimal point (.) followed by a sequence of decimal digits. The exponent consists of (\), a carriage return, or a line feed.
an upper- or lowercase e prefix followed by a sequence of decimal digits that indicates
what power of 10 the value preceding the e is multiplied by. For example, 1.25e2 Special characters can be included in string literals using the following escape
represents 1.25 x 102, which evaluates to 125.0. Similarly, 1.25e-2 represents 1.25 x sequences:
10-2, which evaluates to 0.0125.
Null Character (\0)
Hexadecimal floating-point literals consist of a 0x prefix, followed by an optional
hexadecimal fraction, followed by a hexadecimal exponent. The hexadecimal fraction Backslash (\\)
consists of a decimal point followed by a sequence of hexadecimal digits. The Horizontal Tab (\t)

190
 Line Feed (\n) The Swift standard library defines a number of operators for your use, many of which
are discussed in Basic Operators and Advanced Operators. The present section
Carriage Return (\r)
describes which characters can be used to define custom operators.
Double Quote (\")
Single Quote (\') Custom operators can begin with one of the ASCII characters /, =, -, +, !, *, %, <, >, &,
|, ^, ?, or ~, or one of the Unicode characters defined in the grammar below (which
Unicode scalar (\u{n}), where n is between one and eight hexadecimal include characters from the Mathematical Operators, Miscellaneous Symbols, and
digits Dingbats Unicode blocks, among others). After the first character, combining Unicode
The value of an expression can be inserted into a string literal by placing the characters are also allowed.
expression in parentheses after a backslash (\). The interpolated expression can
contain a string literal, but can’t contain an unescaped backslash (\), a carriage return, You can also define custom operators that begin with a dot (.). These operators can
or a line feed. contain additional dots. For example, .+. is treated as a single operator. If an operator
doesn’t begin with a dot, it can’t contain a dot elsewhere. For example, +.+ is treated
For example, all the following string literals have the same value: as the + operator followed by the .+ operator.

"1 2 3" Although you can define custom operators that contain a question mark (?), they can’t
"1 2 \("3")" consist of a single question mark character only. Additionally, although operators can
"1 2 \(3)" contain an exclamation mark (!), postfix operators cannot begin with either a question
mark or an exclamation mark.
"1 2 \(1 + 2)"
let x = 3; "1 2 \(x)" NOTE
The default inferred type of a string literal is String. For more information about the The tokens =, ->, //, /*, */, ., the prefix operators <, &, and ?, the infix operator ?, and
String type, see Strings and Characters and String Structure Reference. the postfix operators >, !, and ? are reserved. These tokens can’t be overloaded, nor
can they be used as custom operators.
String literals that are concatenated by the + operator are concatenated at compile The whitespace around an operator is used to determine whether an operator is used
time. For example, the values of textA and textB in the example below are identical— as a prefix operator, a postfix operator, or a binary operator. This behavior is
no runtime concatenation is performed. summarized in the following rules:
let textA = "Hello " + "world"
If an operator has whitespace around both sides or around neither side, it
let textB = "Hello world" is treated as a binary operator. As an example, the +++ operator in a+++b
GRAMMAR OF A STRING LITERAL and a +++ b is treated as a binary operator.

string-literal → static-string-literal interpolated-string-literal If an operator has whitespace on the left side only, it is treated as a prefix
static-string-literal → "quoted-textopt" unary operator. As an example, the +++ operator in a +++b is treated as a
quoted-text → quoted-text-itemquoted-textopt prefix unary operator.
quoted-text-item → escaped-character If an operator has whitespace on the right side only, it is treated as a postfix
quoted-text-item → Any Unicode scalar value except ", \, U+000A, or U+000D unary operator. As an example, the +++ operator in a+++ b is treated as a
interpolated-string-literal → "interpolated-textopt" postfix unary operator.
interpolated-text → interpolated-text-iteminterpolated-textopt
interpolated-text-item → \(expression) quoted-text-item If an operator has no whitespace on the left but is followed immediately by
escaped-character → \0 \\ \t \n \r \" \' a dot (.), it is treated as a postfix unary operator. As an example, the +++
escaped-character → \u{unicode-scalar-digits} operator in a+++.b is treated as a postfix unary operator (a+++ .b rather
unicode-scalar-digits → Between one and eight hexadecimal digits than a +++ .b).

Operators

191
For the purposes of these rules, the characters (, [, and { before an operator, the operator-character → U+E0100–U+E01EF
characters ), ], and } after an operator, and the characters ,, ;, and : are also operator-characters → operator-characteroperator-charactersopt
considered whitespace. dot-operator-head → .
dot-operator-character → . operator-character
There is one caveat to the rules above. If the ! or ? predefined operator has no dot-operator-characters → dot-operator-characterdot-operator-charactersopt
whitespace on the left, it is treated as a postfix operator, regardless of whether it has binary-operator → operator
whitespace on the right. To use the ? as the optional-chaining operator, it must not prefix-operator → operator
have whitespace on the left. To use it in the ternary conditional (? :) operator, it must postfix-operator → operator
have whitespace around both sides.

In certain constructs, operators with a leading < or > may be split into two or more
tokens. The remainder is treated the same way and may be split again. As a result,
there is no need to use whitespace to disambiguate between the closing > characters
in constructs like Dictionary<String, Array<Int>>. In this example, the closing >
characters are not treated as a single token that may then be misinterpreted as a bit
shift >> operator.

To learn how to define new, custom operators, see Custom Operators and Operator
Declaration. To learn how to overload existing operators, see Operator Methods.

G R A M M A R O F O P E R AT O R S

operator → operator-headoperator-charactersopt
operator → dot-operator-headdot-operator-characters
operator-head → / = - + ! * % < > & | ^ ~ ?
operator-head → U+00A1–U+00A7
operator-head → U+00A9 or U+00AB
operator-head → U+00AC or U+00AE
operator-head → U+00B0–U+00B1, U+00B6, U+00BB, U+00BF, U+00D7, or
U+00F7
operator-head → U+2016–U+2017 or U+2020–U+2027
operator-head → U+2030–U+203E
operator-head → U+2041–U+2053
operator-head → U+2055–U+205E
operator-head → U+2190–U+23FF
operator-head → U+2500–U+2775
operator-head → U+2794–U+2BFF
operator-head → U+2E00–U+2E7F
operator-head → U+3001–U+3003
operator-head → U+3008–U+3030
operator-character → operator-head
operator-character → U+0300–U+036F
operator-character → U+1DC0–U+1DFF
operator-character → U+20D0–U+20FF
operator-character → U+FE00–U+FE0F
operator-character → U+FE20–U+FE2F

192
 Section 3 In the first example, the expression someTuple is specified to have the tuple type
(Double, Double). In the second example, the parameter a to the function

Types someFunction is specified to have the type Int.

Type annotations can contain an optional list of type attributes before the type.

G R A M M A R O F A T Y P E A N N O TAT I O N

Types type-annotation → :attributesoptinoutopttype

In Swift, there are two kinds of types: named types and compound types. A named Type Identifier
type is a type that can be given a particular name when it is defined. Named types
include classes, structures, enumerations, and protocols. For example, instances of a A type identifier refers to either a named type or a type alias of a named or compound
user-defined class named MyClass have the type MyClass. In addition to user-defined type.
named types, the Swift standard library defines many commonly used named types,
including those that represent arrays, dictionaries, and optional values. Most of the time, a type identifier directly refers to a named type with the same name
as the identifier. For example, Int is a type identifier that directly refers to the named
Data types that are normally considered basic or primitive in other languages—such type Int, and the type identifier Dictionary<String, Int> directly refers to the
as types that represent numbers, characters, and strings—are actually named types, named type Dictionary<String, Int>.
defined and implemented in the Swift standard library using structures. Because they
are named types, you can extend their behavior to suit the needs of your program, There are two cases in which a type identifier does not refer to a type with the same
using an extension declaration, discussed in Extensions and Extension Declaration. name. In the first case, a type identifier refers to a type alias of a named or compound
type. For instance, in the example below, the use of Point in the type annotation
A compound type is a type without a name, defined in the Swift language itself. There refers to the tuple type (Int, Int).
are two compound types: function types and tuple types. A compound type may
typealias Point = (Int, Int)
contain named types and other compound types. For instance, the tuple type (Int,
(Int, Int)) contains two elements: The first is the named type Int, and the second let origin: Point = (0, 0)
is another compound type (Int, Int). In the second case, a type identifier uses dot (.) syntax to refer to named types
declared in other modules or nested within other types. For example, the type
This chapter discusses the types defined in the Swift language itself and describes identifier in the following code references the named type MyType that is declared in
the type inference behavior of Swift. the ExampleModule module.

GRAMMAR OF A TYPE var someValue: ExampleModule.MyType

type → array-type dictionary-type function-type type-identifier tuple-type optional- GRAMMAR OF A TYPE IDENTIFIER
type implicitly-unwrapped-optional-type protocol-composition-type metatype-
type-identifier → type-namegeneric-argument-clauseopt type-namegeneric-
type Any Self
argument-clauseopt.type-identifier
type-name → identifier
Type Annotation
A type annotation explicitly specifies the type of a variable or expression. Type Tuple Type
annotations begin with a colon (:) and end with a type, as the following examples
A tuple type is a comma-separated list of types, enclosed in parentheses.
show:

let someTuple: (Double, Double) = (3.14159, 2.71828) You can use a tuple type as the return type of a function to enable the function to
return a single tuple containing multiple values. You can also name the elements of a
func someFunction(a: Int) { /* ... */ }
tuple type and use those names to refer to the values of the individual elements. An
element name consists of an identifier followed immediately by a colon (:). For an

193
example that demonstrates both of these features, see Functions with Multiple Return Argument names in functions and methods are not part of the corresponding function
Values. type. For example:

When an element of a tuple type has a name, that name is part of the type. func someFunction(left: Int, right: Int) {}
func anotherFunction(left: Int, right: Int) {}
var someTuple = (top: 10, bottom: 12) // someTuple is of type
func functionWithDifferentLabels(top: Int, bottom: Int) {}
(top: Int, bottom: Int)
someTuple = (top: 4, bottom: 42) // OK: names match
var f = someFunction // The type of f is (Int, Int) -> Void, not
someTuple = (9, 99) // OK: names are inferred
(left: Int, right: Int) -> Void.
someTuple = (left: 5, right: 5) // Error: names don't match
f = anotherFunction // OK
All tuple types contain two or more types, except for Void which is a type alias for the f = functionWithDifferentLabels // OK
empty tuple type, (). A single parenthesized type is the same as that type without
parentheses. For example, (Int) is equivalent to Int.
func functionWithDifferentArgumentTypes(left: Int, right: String)
{}
GRAMMAR OF A TUPLE TYPE
func functionWithDifferentNumberOfArguments(left: Int, right: Int,
tuple-type → (tuple-type-element-listopt) top: Int) {}
tuple-type-element-list → tuple-type-element tuple-type-element,tuple-type-
element-list f = functionWithDifferentArgumentTypes // Error
tuple-type-element → element-nametype-annotation type
f = functionWithDifferentNumberOfArguments // Error
element-name → identifier
If a function type includes more than a single arrow (->), the function types are
Function Type grouped from right to left. For example, the function type (Int) -> (Int) -> Int is
understood as (Int) -> ((Int) -> Int)—that is, a function that takes an Int and
A function type represents the type of a function, method, or closure and consists of a
returns another function that takes and returns an Int.
parameter and return type separated by an arrow (->):

(parameter type) -> return type Function types that can throw an error must be marked with the throws keyword, and
function types that can rethrow an error must be marked with the rethrows keyword.
The parameter type is comma-separated list of types. Because the return type can be
The throws keyword is part of a function’s type, and nonthrowing functions are
a tuple type, function types support functions and methods that return multiple values.
subtypes of throwing functions. As a result, you can use a nonthrowing function in the
same places as a throwing one. Throwing and rethrowing functions are described in
A parameter of the function type () -> T (where T is any type) can apply the
Throwing Functions and Methods and Rethrowing Functions and Methods.
autoclosure attribute to implicitly create a closure at its call sites. This provides a
syntactically convenient way to defer the evaluation of an expression without needing GRAMMAR OF A FUNCTION TYPE
to write an explicit closure when you call the function. For an example of an
autoclosure function type parameter, see Autoclosures. function-type → attributesoptfunction-type-argument-clausethrowsopt->type
function-type → attributesoptfunction-type-argument-clauserethrows->type
A function type can have a variadic parameter in its parameter type. Syntactically, a function-type-argument-clause → ()
variadic parameter consists of a base type name followed immediately by three dots function-type-argument-clause → (function-type-argument-list...opt)
(...), as in Int.... A variadic parameter is treated as an array that contains elements function-type-argument-list → function-type-argument function-type-
of the base type name. For instance, the variadic parameter Int... is treated as argument,function-type-argument-list
[Int]. For an example that uses a variadic parameter, see Variadic Parameters. function-type-argument → attributesoptinoutopttype argument-labeltype-annotation
argument-label → identifier
To specify an in-out parameter, prefix the parameter type with the inout keyword. You
can’t mark a variadic parameter or a return type with the inout keyword. In-out Array Type
parameters are discussed in In-Out Parameters.

194
The Swift language provides the following syntactic sugar for the Swift standard library The values of a dictionary can be accessed through subscripting by specifying the
Array<Element> type: corresponding key in square brackets: someDictionary["Alex"] refers to the value
associated with the key "Alex". The subscript returns an optional value of the
[type] dictionary’s value type. If the specified key isn’t contained in the dictionary, the
In other words, the following two declarations are equivalent: subscript returns nil.

let someArray: Array<String> = ["Alex", "Brian", "Dave"] The key type of a dictionary must conform to the Swift standard library Hashable
let someArray: [String] = ["Alex", "Brian", "Dave"] protocol.
In both cases, the constant someArray is declared as an array of strings. The elements
For a detailed discussion of the Swift standard library Dictionary type, see
of an array can be accessed through subscripting by specifying a valid index value in
Dictionaries.
square brackets: someArray[0] refers to the element at index 0, "Alex".
GRAMMAR OF A DICTIONARY TYPE
You can create multidimensional arrays by nesting pairs of square brackets, where the
name of the base type of the elements is contained in the innermost pair of square dictionary-type → [type:type]
brackets. For example, you can create a three-dimensional array of integers using
three sets of square brackets: Optional Type
The Swift language defines the postfix ? as syntactic sugar for the named type
var array3D: [[[Int]]] = [[[1, 2], [3, 4]], [[5, 6], [7, 8]]]
Optional<Wrapped>, which is defined in the Swift standard library. In other words, the
When accessing the elements in a multidimensional array, the left-most subscript following two declarations are equivalent:
index refers to the element at that index in the outermost array. The next subscript
index to the right refers to the element at that index in the array that’s nested one level var optionalInteger: Int?
in. And so on. This means that in the example above, array3D[0] refers to [[1, 2], var optionalInteger: Optional<Int>
[3, 4]], array3D[0][1] refers to [3, 4], and array3D[0][1][1] refers to the value 4.
In both cases, the variable optionalInteger is declared to have the type of an
optional integer. Note that no whitespace may appear between the type and the ?.
For a detailed discussion of the Swift standard library Array type, see Arrays.

G R A M M A R O F A N A R R AY T Y P E The type Optional<Wrapped> is an enumeration with two cases, none and

some(Wrapped), which are used to represent values that may or may not be present.
array-type → [type] Any type can be explicitly declared to be (or implicitly converted to) an optional type. If
you don’t provide an initial value when you declare an optional variable or property, its
Dictionary Type value automatically defaults to nil.
The Swift language provides the following syntactic sugar for the Swift standard library
Dictionary<Key, Value> type: If an instance of an optional type contains a value, you can access that value using
the postfix operator !, as shown below:
[key type: value type]
optionalInteger = 42
In other words, the following two declarations are equivalent:
optionalInteger! // 42
let someDictionary: [String: Int] = ["Alex": 31, "Paul": 39] Using the ! operator to unwrap an optional that has a value of nil results in a runtime
let someDictionary: Dictionary<String, Int> = ["Alex": 31, "Paul": error.
39]

In both cases, the constant someDictionary is declared as a dictionary with strings as You can also use optional chaining and optional binding to conditionally perform an
keys and integers as values. operation on an optional expression. If the value is nil, no operation is performed and
therefore no runtime error is produced.

195
For more information and to see examples that show how to use optional types, see G R A M M A R O F A N I M P L I C I T LY U N W R A P P E D O P T I O N A L T Y P E
Optionals. implicitly-unwrapped-optional-type → type!
GRAMMAR OF AN OPTIONAL TYPE
Protocol Composition Type
optional-type → type? A protocol composition type describes a type that conforms to each protocol in a list of
specified protocols. Protocol composition types may be used only in type annotations
Implicitly Unwrapped Optional Type and in generic parameters.
The Swift language defines the postfix ! as syntactic sugar for the named type
Optional<Wrapped>, which is defined in the Swift standard library, with the additional Protocol composition types have the following form:
behavior that it’s automatically unwrapped when it’s accessed. If you try to use an
implicitly unwrapped optional that has a value of nil, you’ll get a runtime error. With Protocol 1 & Protocol 2
the exception of the implicit unwrapping behavior, the following two declarations are A protocol composition type allows you to specify a value whose type conforms to the
equivalent: requirements of multiple protocols without having to explicitly define a new, named
protocol that inherits from each protocol you want the type to conform to. For example,
var implicitlyUnwrappedString: String! specifying a protocol composition type ProtocolA & ProtocolB & ProtocolC is
var explicitlyUnwrappedString: Optional<String> effectively the same as defining a new protocol ProtocolD that inherits from
ProtocolA, ProtocolB, and ProtocolC, but without having to introduce a new name.
Note that no whitespace may appear between the type and the !.

Each item in a protocol composition list must be either the name of protocol or a type
Because implicit unwrapping changes the meaning of the declaration that contains
alias of a protocol composition type.
that type, optional types that are nested inside a tuple type or a generic type—such as
the element types of a dictionary or array—can’t be marked as implicitly unwrapped.
GRAMMAR OF A PROTOCOL COMPOSITION TYPE
For example:
protocol-composition-type → protocol-identifier&protocol-composition-continuation
let tupleOfImplicitlyUnwrappedElements: (Int!, Int!) // Error protocol-composition-continuation → protocol-identifier protocol-composition-type
let implicitlyUnwrappedTuple: (Int, Int)! // OK protocol-identifier → type-identifier

let arrayOfImplicitlyUnwrappedElements: [Int!] // Error Metatype Type

let implicitlyUnwrappedArray: [Int]! // OK A metatype type refers to the type of any type, including class types, structure types,
enumeration types, and protocol types.
Because implicitly unwrapped optionals have the same Optional<Wrapped> type as
optional values, you can use implicitly unwrapped optionals in all the same places in The metatype of a class, structure, or enumeration type is the name of that type
your code that you can use optionals. For instance, you can assign values of implicitly followed by .Type. The metatype of a protocol type—not the concrete type that
unwrapped optionals to variables, constants, and properties of optionals, and vice conforms to the protocol at runtime—is the name of that protocol followed
versa. by .Protocol. For example, the metatype of the class type SomeClass is
SomeClass.Type and the metatype of the protocol SomeProtocol is
As with optionals, if you don’t provide an initial value when you declare an implicitly SomeProtocol.Protocol.
unwrapped optional variable or property, its value automatically defaults to nil.
You can use the postfix self expression to access a type as a value. For example,
Use optional chaining to conditionally perform an operation on an implicitly unwrapped SomeClass.self returns SomeClass itself, not an instance of SomeClass. And
optional expression. If the value is nil, no operation is performed and therefore no SomeProtocol.self returns SomeProtocol itself, not an instance of a type that
runtime error is produced. conforms to SomeProtocol at runtime. You can use a type(of:) expression with an
instance of a type to access that instance’s dynamic, runtime type as a value, as the
For more information about implicitly unwrapped optional types, see Implicitly following example shows:
Unwrapped Optionals.

196
 class SomeBaseClass { let metatype: AnotherSubClass.Type = AnotherSubClass.self
class func printClassName() { let anotherInstance = metatype.init(string: "some string")
print("SomeBaseClass") G R A M M A R O F A M E TAT Y P E T Y P E
}
metatype-type → type.Type type.Protocol
}
class SomeSubClass: SomeBaseClass { Type Inheritance Clause
override class func printClassName() {
A type inheritance clause is used to specify which class a named type inherits from
print("SomeSubClass") and which protocols a named type conforms to. A type inheritance clause is also used
} to specify a class requirement on a protocol. A type inheritance clause begins with a
} colon (:), followed by either a class requirement, a list of type identifiers, or both.
let someInstance: SomeBaseClass = SomeSubClass()
Class types can inherit from a single superclass and conform to any number of
// The compile-time type of someInstance is SomeBaseClass,
protocols. When defining a class, the name of the superclass must appear first in the
// and the runtime type of someInstance is SomeSubClass list of type identifiers, followed by any number of protocols the class must conform to.
type(of: someInstance).printClassName() If the class does not inherit from another class, the list can begin with a protocol
// Prints "SomeSubClass" instead. For an extended discussion and several examples of class inheritance, see
Inheritance.
Use the identity operators (=== and !==) to test whether an instance’s runtime type is
the same as its compile-time type. Other named types can only inherit from or conform to a list of protocols. Protocol
types can inherit from any number of other protocols. When a protocol type inherits
if type(of: someInstance) === someInstance.self {
from other protocols, the set of requirements from those other protocols are
print("The dynamic and static type of someInstance are the aggregated together, and any type that inherits from the current protocol must conform
same")
to all of those requirements. As discussed in Protocol Declaration, you can include the
} else { class keyword as the first item in the type inheritance clause to mark a protocol
print("The dynamic and static type of someInstance are declaration with a class requirement.
different")
} A type inheritance clause in an enumeration definition can be either a list of protocols,
// Prints "The dynamic and static type of someInstance are or in the case of an enumeration that assigns raw values to its cases, a single, named
different" type that specifies the type of those raw values. For an example of an enumeration
Use an initializer expression to construct an instance of a type from that type’s definition that uses a type inheritance clause to specify the type of its raw values, see
metatype value. For class instances, the initializer that’s called must be marked with Raw Values.
the required keyword or the entire class marked with the final keyword.
G R A M M A R O F A T Y P E I N H E R I TA N C E C L A U S E
class AnotherSubClass: SomeBaseClass { type-inheritance-clause → :class-requirement,type-inheritance-list
let string: String type-inheritance-clause → :class-requirement
required init(string: String) { type-inheritance-clause → :type-inheritance-list
self.string = string
type-inheritance-list → type-identifier type-identifier,type-inheritance-list
class-requirement → class
}
override class func printClassName() {
Type Inference
print("AnotherSubClass")
Swift uses type inference extensively, allowing you to omit the type or part of the type
} of many variables and expressions in your code. For example, instead of writing var
} x: Int = 0, you can write var x = 0, omitting the type completely—the compiler

197
correctly infers that x names a value of type Int. Similarly, you can omit part of a type
when the full type can be inferred from context. For instance, if you write let dict:
Dictionary = ["A": 1], the compiler infers that dict has the type
Dictionary<String, Int>.

In both of the examples above, the type information is passed up from the leaves of
the expression tree to its root. That is, the type of x in var x: Int = 0 is inferred by
first checking the type of 0 and then passing this type information up to the root (the
variable x).

In Swift, type information can also flow in the opposite direction—from the root down
to the leaves. In the following example, for instance, the explicit type annotation (:
Float) on the constant eFloat causes the numeric literal 2.71828 to have an inferred
type of Float instead of Double.

let e = 2.71828 // The type of e is inferred to be Double.

let eFloat: Float = 2.71828 // The type of eFloat is Float.

Type inference in Swift operates at the level of a single expression or statement. This
means that all of the information needed to infer an omitted type or part of a type in an
expression must be accessible from type-checking the expression or one of its
subexpressions.

198
 Section 4 A try expression consists of the try operator followed by an expression that can throw
an error. It has the following form:

Expressions try expression

An optional-try expression consists of the try? operator followed by an expression
that can throw an error. It has the following form:

Expressions try? expression

If the expression does not throw an error, the value of the optional-try expression is an
In Swift, there are four kinds of expressions: prefix expressions, binary expressions, optional containing the value of the expression. Otherwise, the value of the optional-
primary expressions, and postfix expressions. Evaluating an expression returns a try expression is nil.
value, causes a side effect, or both.
A forced-try expression consists of the try! operator followed by an expression that
Prefix and binary expressions let you apply operators to smaller expressions. Primary can throw an error. It has the following form:
expressions are conceptually the simplest kind of expression, and they provide a way
to access values. Postfix expressions, like prefix and binary expressions, let you build try! expression
up more complex expressions using postfixes such as function calls and member If the expression throws an error, a runtime error is produced.
access. Each kind of expression is described in detail in the sections below.
When the expression on the left hand side of a binary operator is marked with try,
GRAMMAR OF AN EXPRESSION try?, or try!, that operator applies to the whole binary expression. That said, you can
use parentheses to be explicit about the scope of the operator’s application.
expression → try-operatoroptprefix-expressionbinary-expressionsopt
expression-list → expression expression,expression-list sum = try someThrowingFunction() + anotherThrowingFunction() //
try applies to both function calls
Prefix Expressions sum = try (someThrowingFunction() + anotherThrowingFunction()) //
Prefix expressions combine an optional prefix operator with an expression. Prefix try applies to both function calls
operators take one argument, the expression that follows them. sum = (try someThrowingFunction()) + anotherThrowingFunction() //
Error: try applies only to the first function call

For information about the behavior of these operators, see Basic Operators and A try expression can’t appear on the right hand side of a binary operator, unless the
Advanced Operators. binary operator is the assignment operator or the try expression is enclosed in
parentheses.
For information about the operators provided by the Swift standard library, see Swift
Standard Library Operators Reference. For more information and to see examples of how to use try, try?, and try!, see
Error Handling.
In addition to the standard library operators, you use & immediately before the name
of a variable that’s being passed as an in-out argument to a function call expression. GRAMMAR OF A TRY EXPRESSION
For more information and to see an example, see In-Out Parameters. try-operator → try try? try!
GRAMMAR OF A PREFIX EXPRESSION
Binary Expressions
prefix-expression → prefix-operatoroptpostfix-expression
Binary expressions combine an infix binary operator with the expression that it takes
prefix-expression → in-out-expression
as its left-hand and right-hand arguments. It has the following form:
in-out-expression → &identifier
left-hand argument operator right-hand argument
Try Operator

199
For information about the behavior of these operators, see Basic Operators and For an example that uses the ternary conditional operator, see Ternary Conditional
Advanced Operators. Operator.

For information about the operators provided by the Swift standard library, see Swift G R A M M A R O F A C O N D I T I O N A L O P E R AT O R
Standard Library Operators Reference. conditional-operator → ?try-operatoroptexpression:
NOTE
Type-Casting Operators
At parse time, an expression made up of binary operators is represented as a flat list. There are four type-casting operators: the is operator, the as operator, the as?
This list is transformed into a tree by applying operator precedence. For example, the operator, and the as! operator.
expression 2 + 3 * 5 is initially understood as a flat list of five items, 2, +, 3, *, and 5.
This process transforms it into the tree (2 + (3 * 5)).
They have the following form:
GRAMMAR OF A BINARY EXPRESSION
expression is type
binary-expression → binary-operatorprefix-expression expression as type
binary-expression → assignment-operatortry-operatoroptprefix-expression expression as? type
binary-expression → conditional-operatortry-operatoroptprefix-expression expression as! type
binary-expression → type-casting-operator The is operator checks at runtime whether the expression can be cast to the specified
binary-expressions → binary-expressionbinary-expressionsopt type. It returns true if the expression can be cast to the specified type; otherwise, it
returns false.
Assignment Operator
The assignment operator sets a new value for a given expression. It has the following The as operator performs a cast when it is known at compile time that the cast always
form: succeeds, such as upcasting or bridging. Upcasting lets you use an expression as an
instance of its type’s supertype, without using an intermediate variable. The following
expression = value approaches are equivalent:
The value of the expression is set to the value obtained by evaluating the value. If the
expression is a tuple, the value must be a tuple with the same number of elements. func f(_ any: Any) { print("Function for Any") }
(Nested tuples are allowed.) Assignment is performed from each part of the value to func f(_ int: Int) { print("Function for Int") }
the corresponding part of the expression. For example: let x = 10
f(x)
(a, _, (b, c)) = ("test", 9.45, (12, 3))
// Prints "Function for Int"
// a is "test", b is 12, c is 3, and 9.45 is ignored

The assignment operator does not return any value.

let y: Any = x
G R A M M A R O F A N A S S I G N M E N T O P E R AT O R f(y)

assignment-operator → = // Prints "Function for Any"

Ternary Conditional Operator f(x as Any)

The ternary conditional operator evaluates to one of two given values based on the // Prints "Function for Any"
value of a condition. It has the following form:
Bridging lets you use an expression of a Swift standard library type such as String as
condition ? expression used if true : expression used if false its corresponding Foundation type such as NSString without needing to create a new
instance. For more information on bridging, see Working with Cocoa Data Types in
If the condition evaluates to true, the conditional operator evaluates the first
Using Swift with Cocoa and Objective-C (Swift 3.0.1).
expression and returns its value. Otherwise, it evaluates the second expression and
returns its value. The unused expression is not evaluated.

200
The as? operator performs a conditional cast of the expression to the specified type.
Literal Type Value
The as? operator returns an optional of the specified type. At runtime, if the cast
succeeds, the value of expression is wrapped in an optional and returned; otherwise, #file String The name of the file in which it appears.
the value returned is nil. If casting to the specified type is guaranteed to fail or is
guaranteed to succeed, a compile-time error is raised. #line Int The line number on which it appears.

The as! operator performs a forced cast of the expression to the specified type. The #column Int The column number in which it begins.
as! operator returns a value of the specified type, not an optional type. If the cast fails,
a runtime error is raised. The behavior of x as! T is the same as the behavior of (x #function String The name of the declaration in which it appears.
as? T)!.
Inside a function, the value of #function is the name of that function, inside a method
For more information about type casting and to see examples that use the type- it is the name of that method, inside a property getter or setter it is the name of that
casting operators, see Type Casting. property, inside special members like init or subscript it is the name of that
keyword, and at the top level of a file it is the name of the current module.
G R A M M A R O F A T Y P E - C A S T I N G O P E R AT O R
When used as the default value of a function or method, the special literal’s value is
type-casting-operator → istype
determined when the default value expression is evaluated at the call site.
type-casting-operator → astype
type-casting-operator → as?type func logFunctionName(string: String = #function) {
type-casting-operator → as!type
print(string)
}
Primary Expressions
func myFunction() {
Primary expressions are the most basic kind of expression. They can be used as
expressions on their own, and they can be combined with other tokens to make prefix logFunctionName() // Prints "myFunction()".
expressions, binary expressions, and postfix expressions. }

An array literal is an ordered collection of values. It has the following form:

GRAMMAR OF A PRIMARY EXPRESSION

primary-expression → identifiergeneric-argument-clauseopt [value 1, value 2, ...]

primary-expression → literal-expression The last expression in the array can be followed by an optional comma. The value of
primary-expression → self-expression an array literal has type [T], where T is the type of the expressions inside it. If there
primary-expression → superclass-expression are expressions of multiple types, T is their closest common supertype. Empty array
primary-expression → closure-expression literals are written using an empty pair of square brackets and can be used to create
primary-expression → parenthesized-expression an empty array of a specified type.
primary-expression → tuple-expression
primary-expression → implicit-member-expression var emptyArray: [Double] = []
primary-expression → wildcard-expression A dictionary literal is an unordered collection of key-value pairs. It has the following
primary-expression → selector-expression form:
primary-expression → key-path-expression
[key 1: value 1, key 2: value 2, ...]
Literal Expression
The last expression in the dictionary can be followed by an optional comma. The value
A literal expression consists of either an ordinary literal (such as a string or a number), of a dictionary literal has type [Key: Value], where Key is the type of its key
an array or dictionary literal, a playground literal, or one of the following special expressions and Value is the type of its value expressions. If there are expressions of
literals: multiple types, Key and Value are the closest common supertype for their respective
values. An empty dictionary literal is written as a colon inside a pair of brackets ([:])

201
to distinguish it from an empty array literal. You can use an empty dictionary literal to var greeting: String
create an empty dictionary literal of specified key and value types. init(greeting: String) {
self.greeting = greeting
var emptyDictionary: [String: Double] = [:]
}
A playground literal is used by Xcode to create an interactive representation of a color, }
file, or image within the program editor. Playground literals in plain text outside of
Xcode are represented using a special literal syntax. In a mutating method of a value type, you can assign a new instance of that value
type to self. For example:
For information on using playground literals in Xcode, see Xcode Help > Use
struct Point {
playgrounds > Add a literal.
var x = 0.0, y = 0.0
GRAMMAR OF A LITERAL EXPRESSION mutating func moveBy(x deltaX: Double, y deltaY: Double) {

literal-expression → literal self = Point(x: x + deltaX, y: y + deltaY)

literal-expression → array-literal dictionary-literal playground-literal }
literal-expression → #file #line #column #function }
array-literal → [array-literal-itemsopt]
array-literal-items → array-literal-item,opt array-literal-item,array-literal-items GRAMMAR OF A SELF EXPRESSION

array-literal-item → expression self-expression → self self-method-expression self-subscript-expression self-

dictionary-literal → [dictionary-literal-items] [:] initializer-expression
dictionary-literal-items → dictionary-literal-item,opt dictionary-literal-item,dictionary- self-method-expression → self.identifier
literal-items self-subscript-expression → self[expression-list]
dictionary-literal-item → expression:expression self-initializer-expression → self.init
playground-literal →
#colorLiteral(red:expression,green:expression,blue:expression,alpha:e Superclass Expression
xpression) A superclass expression lets a class interact with its superclass. It has one of the
playground-literal → #fileLiteral(resourceName:expression) following forms:
playground-literal → #imageLiteral(resourceName:expression)
super.member name
Self Expression super[subscript index]
The self expression is an explicit reference to the current type or instance of the type super.init(initializer arguments)
in which it occurs. It has the following forms: The first form is used to access a member of the superclass. The second form is used
to access the superclass’s subscript implementation. The third form is used to access
self an initializer of the superclass.
self.member name
self[subscript index]
self(initializer arguments) Subclasses can use a superclass expression in their implementation of members,
self.init(initializer arguments) subscripting, and initializers to make use of the implementation in their superclass.
In an initializer, subscript, or instance method, self refers to the current instance of
GRAMMAR OF A SUPERCLASS EXPRESSION
the type in which it occurs. In a type method, self refers to the current type in which it
occurs. superclass-expression → superclass-method-expression superclass-subscript-
expression superclass-initializer-expression
The self expression is used to specify scope when accessing members, providing superclass-method-expression → super.identifier
disambiguation when there is another variable of the same name in scope, such as a superclass-subscript-expression → super[expression-list]
function parameter. For example: superclass-initializer-expression → super.init

class SomeClass {

202
Closure Expression Capture Lists
A closure expression creates a closure, also known as a lambda or an anonymous By default, a closure expression captures constants and variables from its surrounding
function in other programming languages. Like a function declaration, a closure scope with strong references to those values. You can use a capture list to explicitly
contains statements which it executes, and it captures constants and variables from control how values are captured in a closure.
its enclosing scope. It has the following form:
A capture list is written as a comma separated list of expressions surrounded by
{ (parameters) -> return type in square brackets, before the list of parameters. If you use a capture list, you must also
statements
}
use the in keyword, even if you omit the parameter names, parameter types, and
return type.
The parameters have the same form as the parameters in a function declaration, as
described in Function Declaration.
The entries in the capture list are initialized when the closure is created. For each
entry in the capture list, a constant is initialized to the value of the constant or variable
There are several special forms that allow closures to be written more concisely:
that has the same name in the surrounding scope. For example in the code below, a is
included in the capture list but b is not, which gives them different behavior.
A closure can omit the types of its parameters, its return type, or both. If
you omit the parameter names and both types, omit the in keyword before var a = 0
the statements. If the omitted types can’t be inferred, a compile-time error
var b = 0
is raised.
let closure = { [a] in
A closure may omit names for its parameters. Its parameters are then
print(a, b)
implicitly named $ followed by their position: $0, $1, $2, and so on.
}
A closure that consists of only a single expression is understood to return
the value of that expression. The contents of this expression are also
considered when performing type inference on the surrounding expression. a = 10
b = 10
The following closure expressions are equivalent:
closure()
myFunction { // Prints "0 10"
(x: Int, y: Int) -> Int in There are two different things named a, the variable in the surrounding scope and the
return x + y constant in the closure’s scope, but only one variable named b. The a in the inner
} scope is initialized with the value of the a in the outer scope when the closure is
created, but their values are not connected in any special way. This means that a
change to the value of a in the outer scope does not affect the value of a in the inner
myFunction { scope, nor does a change to a inside the closure affect the value of a outside the
(x, y) in closure. In contrast, there is only one variable named b—the b in the outer scope—so
return x + y changes from inside or outside the closure are visible in both places.
}
This distinction is not visible when the captured variable’s type has reference
semantics. For example, there are two things named x in the code below, a variable in
myFunction { return $0 + $1 } the outer scope and a constant in the inner scope, but they both refer to the same
object because of reference semantics.
myFunction { $0 + $1 }
class SimpleClass {
For information about passing a closure as an argument to a function, see Function
var value: Int = 0
Call Expression.
}

203
 var x = SimpleClass() capture-list-item → capture-specifieroptexpression
var y = SimpleClass() capture-specifier → weak unowned unowned(safe) unowned(unsafe)
let closure = { [x] in
Implicit Member Expression
print(x.value, y.value)
An implicit member expression is an abbreviated way to access a member of a type,
}
such as an enumeration case or a type method, in a context where type inference can
determine the implied type. It has the following form:
x.value = 10
y.value = 10 .member name

closure()
For example:
// Prints "10 10" var x = MyEnumeration.someValue
If the type of the expression’s value is a class, you can mark the expression in a x = .anotherValue
capture list with weak or unowned to capture a weak or unowned reference to the
GRAMMAR OF A IMPLICIT MEMBER EXPRESSION
expression’s value.
implicit-member-expression → .identifier
myFunction { print(self.title) } // strong
capture Parenthesized Expression
myFunction { [weak self] in print(self!.title) } // weak capture A parenthesized expression consists of an expression surrounded by parentheses.
myFunction { [unowned self] in print(self.title) } // unowned You can use parentheses to specify the precedence of operations by explicitly
capture grouping expressions. Grouping parentheses don’t change an expression’s type—for
You can also bind an arbitrary expression to a named value in a capture list. The example, the type of (1) is simply Int.
expression is evaluated when the closure is created, and the value is captured with
the specified strength. For example: G R A M M A R O F A PA R E N T H E S I Z E D E X P R E S S I O N

parenthesized-expression → (expression)
// Weak capture of "self.parent" as "parent"
myFunction { [weak parent = self.parent] in print(parent!.title) } Tuple Expression
For more information and examples of closure expressions, see Closure Expressions. A tuple expression consists of a comma-separated list of expressions surrounded by
For more information and examples of capture lists, see Resolving Strong Reference parentheses. Each expression can have an optional identifier before it, separated by a
Cycles for Closures. colon (:). It has the following form:

GRAMMAR OF A CLOSURE EXPRESSION (identifier 1: expression 1, identifier 2: expression 2, ...)

A tuple expression can contain zero expressions, or it can contain two or more
closure-expression → {closure-signatureoptstatementsopt}
expressions. A single expression inside parentheses is a parenthesized expression.
closure-signature → capture-listoptclosure-parameter-clausethrowsoptfunction-
resultoptin
GRAMMAR OF A TUPLE EXPRESSION
closure-signature → capture-listin
closure-parameter-clause → () (closure-parameter-list) identifier-list tuple-expression → () (tuple-element,tuple-element-list)
closure-parameter-list → closure-parameter closure-parameter,closure-parameter- tuple-element-list → tuple-element tuple-element,tuple-element-list
list tuple-element → expression identifier:expression
closure-parameter → closure-parameter-nametype-annotationopt
closure-parameter → closure-parameter-nametype-annotation... Wildcard Expression
closure-parameter-name → identifier A wildcard expression is used to explicitly ignore a value during an assignment. For
capture-list → [capture-list-items] example, in the following assignment 10 is assigned to x and 20 is ignored:
capture-list-items → capture-list-item capture-list-item,capture-list-items

204
 (x, _) = (10, 20) Because a selector is created at compile time, not at runtime, the compiler can check
// x is 10, and 20 is ignored that a method or property exists and that they’re exposed to the Objective-C runtime.
GRAMMAR OF A WILDCARD EXPRESSION
NOTE
wildcard-expression → _ Although the method name and the property name are expressions, they’re never
evaluated.
Selector Expression
A selector expression lets you access the selector used to refer to a method or to a For more information about using selectors in Swift code that interacts with Objective-
property’s getter or setter in Objective-C. C APIs, see Objective-C Selectors in Using Swift with Cocoa and Objective-C (Swift
3.0.1).
#selector(method name)
#selector(getter: property name) GRAMMAR OF A SELECTOR EXPRESSION
#selector(setter: property name)
selector-expression → #selector(expression)
The method name and property name must be a reference to a method or a property
selector-expression → #selector(getter:expression)
that is available in the Objective-C runtime. The value of a selector expression is an
selector-expression → #selector(setter:expression)
instance of the Selector type. For example:
Key-Path Expression
class SomeClass: NSObject {
A key-path expression lets you access the string used to refer to a property in
let property: String
Objective-C for use in key-value coding and key-value observing APIs.
@objc(doSomethingWithInt:)
func doSomething(_ x: Int) {} #keyPath(property name)
The property name must be a reference to a property that is available in the Objective-
init(property: String) {
C runtime. At compile time, the key-path expression is replaced by a string literal. For
example:
self.property = property
} @objc class SomeClass: NSObject {
} var someProperty: Int
let selectorForMethod = #selector(SomeClass.doSomething(_:)) init(someProperty: Int) {
let selectorForPropertyGetter = #selector(getter: self.someProperty = someProperty
SomeClass.property)
}
When creating a selector for a property’s getter, the property name can be a reference func keyPathTest() -> String {
to a variable or constant property. In contrast, when creating a selector for a property’s
return #keyPath(someProperty)
setter, the property name must be a reference to a variable property only.
}
The method name can contain parentheses for grouping, as well the as operator to }
disambiguate between methods that share a name but have different type signatures.
For example: let c = SomeClass(someProperty: 12)
let keyPath = #keyPath(SomeClass.someProperty)
extension SomeClass {
print(keyPath == c.keyPathTest())
@objc(doSomethingWithString:)
// Prints "true"
func doSomething(_ x: String) { }
}
if let value = c.value(forKey: keyPath) {
let anotherSelector = #selector(SomeClass.doSomething(_:) as
(SomeClass) -> (String) -> Void) print(value)

205
 } function name(argument value 1, argument value 2)
// Prints "12" The function name can be any expression whose value is of a function type.
Because the key path is created at compile time, not at runtime, the compiler can
check that the property exists and that the property is exposed to the Objective-C If the function definition includes names for its parameters, the function call must
runtime. include names before its argument values separated by a colon (:). This kind of
function call expression has the following form:
For more information about using selectors in Swift code that interacts with Objective-
function name(argument name 1: argument value 1, argument name 2:
C APIs, see Keys and Key Paths in Using Swift with Cocoa and Objective-C (Swift argument value 2)
3.0.1). For information about key-value coding and key-value observing, see Key-
A function call expression can include a trailing closure in the form of a closure
Value Coding Programming Guide and Key-Value Observing Programming Guide.
expression immediately after the closing parenthesis. The trailing closure is
understood as an argument to the function, added after the last parenthesized
NOTE
argument. The following function calls are equivalent:
Although the property name is an expression, it is never evaluated.
// someFunction takes an integer and a closure as its arguments
G R A M M A R O F A K E Y- PAT H E X P R E S S I O N
someFunction(x: x, f: {$0 == 13})
key-path-expression → #keyPath(expression)
someFunction(x: x) {$0 == 13}

Postfix Expressions If the trailing closure is the function’s only argument, the parentheses can be omitted.

Postfix expressions are formed by applying a postfix operator or other postfix syntax // someFunction takes a closure as its only argument
to an expression. Syntactically, every primary expression is also a postfix expression.
myData.someMethod() {$0 == 13}
myData.someMethod {$0 == 13}
For information about the behavior of these operators, see Basic Operators and
Advanced Operators. GRAMMAR OF A FUNCTION CALL EXPRESSION

function-call-expression → postfix-expressionfunction-call-argument-clause
For information about the operators provided by the Swift standard library, see Swift
function-call-expression → postfix-expressionfunction-call-argument-
Standard Library Operators Reference.
clauseopttrailing-closure
GRAMMAR OF A POSTFIX EXPRESSION
function-call-argument-clause → () (function-call-argument-list)
function-call-argument-list → function-call-argument function-call-
postfix-expression → primary-expression argument,function-call-argument-list
postfix-expression → postfix-expressionpostfix-operator function-call-argument → expression identifier:expression
postfix-expression → function-call-expression function-call-argument → operator identifier:operator
postfix-expression → initializer-expression trailing-closure → closure-expression
postfix-expression → explicit-member-expression
postfix-expression → postfix-self-expression Initializer Expression
postfix-expression → dynamic-type-expression An initializer expression provides access to a type’s initializer. It has the following
postfix-expression → subscript-expression form:
postfix-expression → forced-value-expression
postfix-expression → optional-chaining-expression expression.init(initializer arguments)
You use the initializer expression in a function call expression to initialize a new
Function Call Expression instance of a type. You also use an initializer expression to delegate to the initializer of
A function call expression consists of a function name followed by a comma-separated a superclass.
list of the function’s arguments in parentheses. Function call expressions have the
following form: class SomeSubClass: SomeSuperClass {

206
 override init() { var t = (10, 20, 30)
// subclass initialization goes here t.0 = t.1
super.init() // Now t is (20, 20, 30)
} The members of a module access the top-level declarations of that module.
}

Like a function, an initializer can be used as a value. For example: To distinguish between methods or initializers whose names differ only by the names
of their arguments, include the argument names in parentheses, with each argument
// Type annotation is required because String has multiple name followed by a colon (:). Write an underscore (_) for an argument with no name.
initializers. To distinguish between overloaded methods, use a type annotation. For example:
let initializer: (Int) -> String = String.init
class SomeClass {
let oneTwoThree = [1, 2, 3].map(initializer).reduce("", +)
func someMethod(x: Int, y: Int) {}
print(oneTwoThree)
func someMethod(x: Int, z: Int) {}
// Prints "123"
func overloadedMethod(x: Int, y: Int) {}
If you specify a type by name, you can access the type’s initializer without using an
func overloadedMethod(x: Int, y: Bool) {}
initializer expression. In all other cases, you must use an initializer expression.
}
let s1 = SomeType.init(data: 3) // Valid let instance = SomeClass()
let s2 = SomeType(data: 1) // Also valid
let a = instance.someMethod // Ambiguous
let s3 = type(of: someValue).init(data: 7) // Valid let b = instance.someMethod(x:y:) // Unambiguous
let s4 = type(of: someValue)(data: 5) // Error

GRAMMAR OF AN INITIALIZER EXPRESSION let d = instance.overloadedMethod // Ambiguous

let d = instance.overloadedMethod(x:y:) // Still ambiguous
initializer-expression → postfix-expression.init
initializer-expression → postfix-expression.init(argument-names) let d: (Int, Bool) -> Void = instance.overloadedMethod(x:y:) //
Unambiguous

Explicit Member Expression If a period appears at the beginning of a line, it is understood as part of an explicit
An explicit member expression allows access to the members of a named type, a member expression, not as an implicit member expression. For example, the following
tuple, or a module. It consists of a period (.) between the item and the identifier of its listing shows chained method calls split over several lines:
member.
let x = [10, 3, 20, 15, 4]
expression.member name .sorted()
The members of a named type are named as part of the type’s declaration or .filter { $0 > 5 }
extension. For example: .map { $0 * 100 }

class SomeClass { GRAMMAR OF AN EXPLICIT MEMBER EXPRESSION

var someProperty = 42 explicit-member-expression → postfix-expression.decimal-digits

} explicit-member-expression → postfix-expression.identifiergeneric-argument-
let c = SomeClass() clauseopt
explicit-member-expression → postfix-expression.identifier(argument-names)
let y = c.someProperty // Member access
argument-names → argument-nameargument-namesopt
The members of a tuple are implicitly named using integers in the order they appear, argument-name → identifier:
starting from zero. For example:

207
Postfix Self Expression dynamic-type-expression → type(of:expression)
A postfix self expression consists of an expression or the name of a type,
immediately followed by .self. It has the following forms: Subscript Expression
A subscript expression provides subscript access using the getter and setter of the
expression.self corresponding subscript declaration. It has the following form:
type.self
The first form evaluates to the value of the expression. For example, x.self evaluates expression[index expressions]
to x. To evaluate the value of a subscript expression, the subscript getter for the
expression’s type is called with the index expressions passed as the subscript
The second form evaluates to the value of the type. Use this form to access a type as parameters. To set its value, the subscript setter is called in the same way.
a value. For example, because SomeClass.self evaluates to the SomeClass type itself,
you can pass it to a function or method that accepts a type-level argument. For information about subscript declarations, see Protocol Subscript Declaration.

GRAMMAR OF A SELF EXPRESSION GRAMMAR OF A SUBSCRIPT EXPRESSION

postfix-self-expression → postfix-expression.self subscript-expression → postfix-expression[expression-list]

Dynamic Type Expression Forced-Value Expression

A dynamic type expression consists of an expression within special syntax that A forced-value expression unwraps an optional value that you are certain is not nil. It
resembles a Function Call Expression. It has the following form: has the following form:

type(of: expression) expression!

The expression can’t be the name of a type. The entire type(of:) expression If the value of the expression is not nil, the optional value is unwrapped and returned
evaluates to the value of the runtime type of the expression, as the following example with the corresponding nonoptional type. Otherwise, a runtime error is raised.
shows:
The unwrapped value of a forced-value expression can be modified, either by
class SomeBaseClass { mutating the value itself, or by assigning to one of the value’s members. For example:
class func printClassName() {
print("SomeBaseClass") var x: Int? = 0

} x! += 1

} // x is now 1

class SomeSubClass: SomeBaseClass {

override class func printClassName() { var someDictionary = ["a": [1, 2, 3], "b": [10, 20]]

print("SomeSubClass") someDictionary["a"]![0] = 100

} // someDictionary is now ["b": [10, 20], "a": [100, 2, 3]]

} G R A M M A R O F A F O R C E D - VA L U E E X P R E S S I O N
let someInstance: SomeBaseClass = SomeSubClass() forced-value-expression → postfix-expression!
// someInstance has a static type of SomeBaseClass at compile time,
and Optional-Chaining Expression
// it has a dynamic type of SomeSubClass at runtime An optional-chaining expression provides a simplified syntax for using optional values
type(of: someInstance).printClassName() in postfix expressions. It has the following form:
// Prints "SomeSubClass"
expression?
GRAMMAR OF A DYNAMIC TYPE EXPRESSION

208
The postfix ? operator makes an optional-chaining expression from an expression // someDictionary is now ["b": [10, 20], "a": [42, 2, 3]]
without changing the expression’s value. GRAMMAR OF AN OPTIONAL-CHAINING EXPRESSION

Optional-chaining expressions must appear within a postfix expression, and they optional-chaining-expression → postfix-expression?
cause the postfix expression to be evaluated in a special way. If the value of the
optional-chaining expression is nil, all of the other operations in the postfix
expression are ignored and the entire postfix expression evaluates to nil. If the value
of the optional-chaining expression is not nil, the value of the optional-chaining
expression is unwrapped and used to evaluate the rest of the postfix expression. In
either case, the value of the postfix expression is still of an optional type.

If a postfix expression that contains an optional-chaining expression is nested inside

other postfix expressions, only the outermost expression returns an optional type. In
the example below, when c is not nil, its value is unwrapped and used to
evaluate .property, the value of which is used to evaluate .performAction(). The
entire expression c?.property.performAction() has a value of an optional type.

var c: SomeClass?
var result: Bool? = c?.property.performAction()

The following example shows the behavior of the example above without using
optional chaining.

var result: Bool? = nil

if let unwrappedC = c {
result = unwrappedC.property.performAction()
}

The unwrapped value of an optional-chaining expression can be modified, either by

mutating the value itself, or by assigning to one of the value’s members. If the value of
the optional-chaining expression is nil, the expression on the right hand side of the
assignment operator is not evaluated. For example:

func someFunctionWithSideEffects() -> Int {

return 42 // No actual side effects.
}
var someDictionary = ["a": [1, 2, 3], "b": [10, 20]]

someDictionary["not here"]?[0] = someFunctionWithSideEffects()

// someFunctionWithSideEffects is not evaluated
// someDictionary is still ["b": [10, 20], "a": [1, 2, 3]]

someDictionary["a"]?[0] = someFunctionWithSideEffects()
// someFunctionWithSideEffects is evaluated and returns 42

209
 Section 5 Control flow in a loop statement can be changed by a break statement and a
continue statement and is discussed in Break Statement and Continue Statement

Statements below.

G R A M M A R O F A L O O P S TAT E M E N T

loop-statement → for-in-statement
loop-statement → while-statement
Statements loop-statement → repeat-while-statement

In Swift, there are three kinds of statements: simple statements, compiler control For-In Statement
statements, and control flow statements. Simple statements are the most common A for-in statement allows a block of code to be executed once for each item in a
and consist of either an expression or a declaration. Compiler control statements collection (or any type) that conforms to the Sequence protocol.
allow the program to change aspects of the compiler’s behavior and include a
conditional compilation block and a line control statement. A for-in statement has the following form:

Control flow statements are used to control the flow of execution in a program. There for item in collection {
are several types of control flow statements in Swift, including loop statements, statements
}
branch statements, and control transfer statements. Loop statements allow a block of
code to be executed repeatedly, branch statements allow a certain block of code to be The makeIterator() method is called on the collection expression to obtain a value of
executed only when certain conditions are met, and control transfer statements an iterator type—that is, a type that conforms to the IteratorProtocol protocol. The
provide a way to alter the order in which code is executed. In addition, Swift provides program begins executing a loop by calling the next() method on the iterator. If the
a do statement to introduce scope, and catch and handle errors, and a defer value returned is not nil, it is assigned to the item pattern, the program executes the
statement for running cleanup actions just before the current scope exits. statements, and then continues execution at the beginning of the loop. Otherwise, the
program does not perform assignment or execute the statements, and it is finished
A semicolon (;) can optionally appear after any statement and is used to separate executing the for-in statement.
multiple statements if they appear on the same line.
G R A M M A R O F A F O R - I N S TAT E M E N T
G R A M M A R O F A S TAT E M E N T for-in-statement → forcaseoptpatterninexpressionwhere-clauseoptcode-block
statement → expression;opt
statement → declaration;opt While Statement
statement → loop-statement;opt A while statement allows a block of code to be executed repeatedly, as long as a
statement → branch-statement;opt condition remains true.
statement → labeled-statement;opt
statement → control-transfer-statement;opt A while statement has the following form:
statement → defer-statement;opt
statement → do-statement:opt while condition {
statements
statement → compiler-control-statement }
statements → statementstatementsopt
A while statement is executed as follows:

Loop Statements 1.The condition is evaluated.

Loop statements allow a block of code to be executed repeatedly, depending on the
If true, execution continues to step 2. If false, the program is finished
conditions specified in the loop. Swift has three loop statements: a for-in statement,
executing the while statement.
a while statement, and a repeat-while statement.
2.The program executes the statements, and execution returns to step 1.

210
Because the value of the condition is evaluated before the statements are executed, statement control how the program branches and, therefore, what block of code is
the statements in a while statement can be executed zero or more times. executed. Swift has three branch statements: an if statement, a guard statement, and
a switch statement.
The value of the condition must be of type Bool or a type bridged to Bool. The
condition can also be an optional binding declaration, as discussed in Optional Control flow in an if statement or a switch statement can be changed by a break
Binding. statement and is discussed in Break Statement below.

G R A M M A R O F A W H I L E S TAT E M E N T G R A M M A R O F A B R A N C H S TAT E M E N T

while-statement → whilecondition-listcode-block branch-statement → if-statement

condition-list → condition condition,condition-list branch-statement → guard-statement
condition → expression availability-condition case-condition optional-binding- branch-statement → switch-statement
condition
case-condition → casepatterninitializer If Statement
optional-binding-condition → letpatterninitializer varpatterninitializer An if statement is used for executing code based on the evaluation of one or more
conditions.
Repeat-While Statement
A repeat-while statement allows a block of code to be executed one or more times, There are two basic forms of an if statement. In each form, the opening and closing
as long as a condition remains true. braces are required.

A repeat-while statement has the following form: The first form allows code to be executed only when a condition is true and has the
following form:
repeat {
statements if condition {
} while condition statements
A repeat-while statement is executed as follows: }
The second form of an if statement provides an additional else clause (introduced by
1.The program executes the statements, and execution continues to step 2. the else keyword) and is used for executing one part of code when the condition is
true and another part of code when the same condition is false. When a single else
2.The condition is evaluated.
clause is present, an if statement has the following form:
If true, execution returns to step 1. If false, the program is finished executing
the repeat-while statement. if condition {
statements to execute if condition is true
Because the value of the condition is evaluated after the statements are executed, the } else {
statements in a repeat-while statement are executed at least once. statements to execute if condition is false
}
The value of the condition must be of type Bool or a type bridged to Bool. The The else clause of an if statement can contain another if statement to test more
condition can also be an optional binding declaration, as discussed in Optional than one condition. An if statement chained together in this way has the following
Binding. form:

G R A M M A R O F A R E P E AT- W H I L E S TAT E M E N T if condition 1 {

statements to execute if condition 1 is true
repeat-while-statement → repeatcode-blockwhileexpression } else if condition 2 {
statements to execute if condition 2 is true
} else {
Branch Statements statements to execute if both conditions are false
}
Branch statements allow the program to execute certain parts of code depending on
the value of one or more conditions. The values of the conditions specified in a branch

211
The value of any condition in an if statement must be of type Bool or a type bridged A switch statement has the following form:
to Bool. The condition can also be an optional binding declaration, as discussed in
Optional Binding. switch control expression {
case pattern 1:
statements
G R A M M A R O F A N I F S TAT E M E N T
case pattern 2 where condition:
statements
if-statement → ifcondition-listcode-blockelse-clauseopt case pattern 3 where condition,
else-clause → elsecode-block elseif-statement pattern 4 where condition:
statements
Guard Statement default:
statements
A guard statement is used to transfer program control out of a scope if one or more }
conditions aren’t met.
The control expression of the switch statement is evaluated and then compared with
the patterns specified in each case. If a match is found, the program executes the
A guard statement has the following form: statements listed within the scope of that case. The scope of each case can’t be
empty. As a result, you must include at least one statement following the colon (:) of
guard condition else {
statements each case label. Use a single break statement if you don’t intend to execute any code
} in the body of a matched case.
The value of any condition in a guard statement must be of type Bool or a type
bridged to Bool. The condition can also be an optional binding declaration, as The values of expressions your code can branch on are very flexible. For instance, in
discussed in Optional Binding. addition to the values of scalar types, such as integers and characters, your code can
branch on the values of any type, including floating-point numbers, strings, tuples,
Any constants or variables assigned a value from an optional binding declaration in a instances of custom classes, and optionals. The value of the control expression can
guard statement condition can be used for the rest of the guard statement’s enclosing even be matched to the value of a case in an enumeration and checked for inclusion
scope. in a specified range of values. For examples of how to use these various types of
values in switch statements, see Switch in Control Flow.
The else clause of a guard statement is required, and must either call a function with
the Never return type or transfer program control outside the guard statement’s A switch case can optionally contain a where clause after each pattern. A where
enclosing scope using one of the following statements: clause is introduced by the where keyword followed by an expression, and is used to
provide an additional condition before a pattern in a case is considered matched to the
return control expression. If a where clause is present, the statements within the relevant
case are executed only if the value of the control expression matches one of the
break
patterns of the case and the expression of the where clause evaluates to true. For
continue
instance, a control expression matches the case in the example below only if it is a
throw tuple that contains two elements of the same value, such as (1, 1).
Control transfer statements are discussed in Control Transfer Statements below. For
case let (x, y) where x == y:
more information on functions with the Never return type, see Functions that Never
Return. As the above example shows, patterns in a case can also bind constants using the
let keyword (they can also bind variables using the var keyword). These constants
G R A M M A R O F A G U A R D S TAT E M E N T (or variables) can then be referenced in a corresponding where clause and throughout
guard-statement → guardcondition-listelsecode-block the rest of the code within the scope of the case. If the case contains multiple patterns
that match the control expression, all of the patterns must contain the same constant
Switch Statement or variable bindings, and each bound variable or constant must have the same type in
all of the case’s patterns.
A switch statement allows certain blocks of code to be executed depending on the
value of a control expression.

212
A switch statement can also include a default case, introduced by the default The scope of a labeled statement is the entire statement following the statement label.
keyword. The code within a default case is executed only if no other cases match the You can nest labeled statements, but the name of each statement label must be
control expression. A switch statement can include only one default case, which must unique.
appear at the end of the switch statement.
For more information and to see examples of how to use statement labels, see
Although the actual execution order of pattern-matching operations, and in particular Labeled Statements in Control Flow.
the evaluation order of patterns in cases, is unspecified, pattern matching in a switch
statement behaves as if the evaluation is performed in source order—that is, the order G R A M M A R O F A L A B E L E D S TAT E M E N T
in which they appear in source code. As a result, if multiple cases contain patterns that
labeled-statement → statement-labelloop-statement
evaluate to the same value, and thus can match the value of the control expression,
labeled-statement → statement-labelif-statement
the program executes only the code within the first matching case in source order.
labeled-statement → statement-labelswitch-statement
labeled-statement → statement-labeldo-statement
Switch Statements Must Be Exhaustive statement-label → label-name:
In Swift, every possible value of the control expression’s type must match the value of label-name → identifier
at least one pattern of a case. When this simply isn’t feasible (for instance, when the
control expression’s type is Int), you can include a default case to satisfy the Control Transfer Statements
requirement.
Control transfer statements can change the order in which code in your program is
executed by unconditionally transferring program control from one piece of code to
Execution Does Not Fall Through Cases Implicitly
another. Swift has five control transfer statements: a break statement, a continue
After the code within a matched case has finished executing, the program exits from statement, a fallthrough statement, a return statement, and a throw statement.
the switch statement. Program execution does not continue or “fall through” to the
next case or default case. That said, if you want execution to continue from one case G R A M M A R O F A C O N T R O L T R A N S F E R S TAT E M E N T
to the next, explicitly include a fallthrough statement, which simply consists of the
fallthrough keyword, in the case from which you want execution to continue. For control-transfer-statement → break-statement
more information about the fallthrough statement, see Fallthrough Statement below. control-transfer-statement → continue-statement
control-transfer-statement → fallthrough-statement
G R A M M A R O F A S W I T C H S TAT E M E N T control-transfer-statement → return-statement
control-transfer-statement → throw-statement
switch-statement → switchexpression{switch-casesopt}
switch-cases → switch-caseswitch-casesopt Break Statement
switch-case → case-labelstatements default-labelstatements
A break statement ends program execution of a loop, an if statement, or a switch
case-label → casecase-item-list:
statement. A break statement can consist of only the break keyword, or it can consist
case-item-list → patternwhere-clauseopt patternwhere-clauseopt,case-item-list
of the break keyword followed by the name of a statement label, as shown below.
default-label → default:
where-clause → wherewhere-expression break
where-expression → expression break label name
When a break statement is followed by the name of a statement label, it ends program
Labeled Statement execution of the loop, if statement, or switch statement named by that label.
You can prefix a loop statement, an if statement, a switch statement, or a do
statement with a statement label, which consists of the name of the label followed When a break statement is not followed by the name of a statement label, it ends
immediately by a colon (:). Use statement labels with break and continue statements program execution of the switch statement or the innermost enclosing loop statement
to be explicit about how you want to change control flow in a loop statement or a in which it occurs. You can’t use an unlabeled break statement to break out of an if
switch statement, as discussed in Break Statement and Continue Statement below. statement.

213
In both cases, program control is then transferred to the first line of code following the A fallthrough statement can appear anywhere inside a switch statement, not just as
enclosing loop or switch statement, if any. the last statement of a case block, but it can’t be used in the final case block. It also
cannot transfer control into a case block whose pattern contains value binding
For examples of how to use a break statement, see Break and Labeled Statements in patterns.
Control Flow.
For an example of how to use a fallthrough statement in a switch statement, see
G R A M M A R O F A B R E A K S TAT E M E N T Control Transfer Statements in Control Flow.
break-statement → breaklabel-nameopt
G R A M M A R O F A FA L LT H R O U G H S TAT E M E N T

Continue Statement fallthrough-statement → fallthrough

A continue statement ends program execution of the current iteration of a loop
statement but does not stop execution of the loop statement. A continue statement Return Statement
can consist of only the continue keyword, or it can consist of the continue keyword A return statement occurs in the body of a function or method definition and causes
followed by the name of a statement label, as shown below. program execution to return to the calling function or method. Program execution
continues at the point immediately following the function or method call.
continue
continue label name
A return statement can consist of only the return keyword, or it can consist of the
When a continue statement is followed by the name of a statement label, it ends return keyword followed by an expression, as shown below.
program execution of the current iteration of the loop statement named by that label.
return
When a continue statement is not followed by the name of a statement label, it ends return expression
program execution of the current iteration of the innermost enclosing loop statement in When a return statement is followed by an expression, the value of the expression is
which it occurs. returned to the calling function or method. If the value of the expression does not
match the value of the return type declared in the function or method declaration, the
In both cases, program control is then transferred to the condition of the enclosing expression’s value is converted to the return type before it is returned to the calling
loop statement. function or method.

In a for statement, the increment expression is still evaluated after the continue NOTE
statement is executed, because the increment expression is evaluated after the As described in Failable Initializers, a special form of the return statement (return
execution of the loop’s body. nil) can be used in a failable initializer to indicate initialization failure.

When a return statement is not followed by an expression, it can be used only to

For examples of how to use a continue statement, see Continue and Labeled
return from a function or method that does not return a value (that is, when the return
Statements in Control Flow.
type of the function or method is Void or ()).
G R A M M A R O F A C O N T I N U E S TAT E M E N T
G R A M M A R O F A R E T U R N S TAT E M E N T
continue-statement → continuelabel-nameopt
return-statement → returnexpressionopt
Fallthrough Statement
Throw Statement
A fallthrough statement consists of the fallthrough keyword and occurs only in a
A throw statement occurs in the body of a throwing function or method, or in the body
case block of a switch statement. A fallthrough statement causes program
of a closure expression whose type is marked with the throws keyword.
execution to continue from one case in a switch statement to the next case. Program
execution continues to the next case even if the patterns of the case label do not
A throw statement causes a program to end execution of the current scope and begin
match the value of the switch statement’s control expression.
error propagation to its enclosing scope. The error that’s thrown continues to
propagate until it’s handled by a catch clause of a do statement.

214
A throw statement consists of the throw keyword followed by an expression, as shown G R A M M A R O F A D E F E R S TAT E M E N T
below. defer-statement → defercode-block
throw expression
The value of the expression must have a type that conforms to the Error protocol.
Do Statement
The do statement is used to introduce a new scope and can optionally contain one or
For an example of how to use a throw statement, see Propagating Errors Using more catch clauses, which contain patterns that match against defined error
Throwing Functions in Error Handling. conditions. Variables and constants declared in the scope of a do statement can be
accessed only within that scope.
G R A M M A R O F A T H R O W S TAT E M E N T
A do statement in Swift is similar to curly braces ({}) in C used to delimit a code block,
throw-statement → throwexpression and does not incur a performance cost at runtime.

Defer Statement A do statement has the following form:

A defer statement is used for executing code just before transferring program control
outside of the scope that the defer statement appears in. do {
try expression
statements
A defer statement has the following form: } catch pattern 1 {
statements
defer { } catch pattern 2 where condition {
statements statements
} }

The statements within the defer statement are executed no matter how program Like a switch statement, the compiler attempts to infer whether catch clauses are
control is transferred. This means that a defer statement can be used, for example, to exhaustive. If such a determination can be made, the error is considered handled.
perform manual resource management such as closing file descriptors, and to perform Otherwise, the error can propagate out of the containing scope, which means the error
actions that need to happen even if an error is thrown. must be handled by an enclosing catch clause or the containing function must be
declared with throws.
If multiple defer statements appear in the same scope, the order they appear is the
reverse of the order they are executed. Executing the last defer statement in a given To ensure that an error is handled, use a catch clause with a pattern that matches all
scope first means that statements inside that last defer statement can refer to errors, such as a wildcard pattern (_). If a catch clause does not specify a pattern, the
resources that will be cleaned up by other defer statements. catch clause matches and binds any error to a local constant named error. For more
information about the pattens you can use in a catch clause, see Patterns.
func f() {
defer { print("First") } To see an example of how to use a do statement with several catch clauses, see
Handling Errors.
defer { print("Second") }
defer { print("Third") } G R A M M A R O F A D O S TAT E M E N T
}
do-statement → docode-blockcatch-clausesopt
f() catch-clauses → catch-clausecatch-clausesopt
// Prints "Third" catch-clause → catchpatternoptwhere-clauseoptcode-block
// Prints "Second"
// Prints "First" Compiler Control Statements
The statements in the defer statement can’t transfer program control outside of the Compiler control statements allow the program to change aspects of the compiler’s
defer statement. behavior. Swift has two complier control statements: a conditional compilation block
and a line control statement.

215
 G R A M M A R O F A C O M P I L E R C O N T R O L S TAT E M E N T #elseif clauses. You can also add a final additional branch using an #else clause.
compiler-control-statement → conditional-compilation-block Conditional compilation blocks that contain multiple branches have the following form:
compiler-control-statement → line-control-statement
#if compilation condition 1
statements to compile if compilation condition 1 is true
Conditional Compilation Block #elseif compilation condition 2
A conditional compilation block allows code to be conditionally compiled depending on statements to compile if compilation condition 2 is true
#else
the value of one or more compilation conditions. statements to compile if both compilation conditions are false
#endif
Every conditional compilation block begins with the #if compilation directive and ends NOTE
with the #endif compilation directive. A simple conditional compilation block has the Each statement in the body of a conditional compilation block is parsed even if it’s not
following form: compiled. However, there is an exception if the compilation condition includes a
swift() platform condition: The statements are parsed only if the compiler’s version of
#if compilation condition Swift matches what is specified in the platform condition. This exception ensures that
statements an older compiler doesn’t attempt to parse syntax introduced in a newer version of
#endif Swift.
Unlike the condition of an if statement, the compilation condition is evaluated at G R A M M A R O F A C O N D I T I O N A L C O M P I L AT I O N B L O C K
compile time. As a result, the statements are compiled and executed only if the
compilation condition evaluates to true at compile time. conditional-compilation-block → if-directive-clauseelseif-directive-clausesoptelse-
directive-clauseoptendif-directive
The compilation condition can include the true and false Boolean literals, an if-directive-clause → if-directivecompilation-conditionstatementsopt
identifier used with the -D command line flag, or any of the platform conditions listed in elseif-directive-clauses → elseif-directive-clauseelseif-directive-clausesopt
the table below. elseif-directive-clause → elseif-directivecompilation-conditionstatementsopt
else-directive-clause → else-directivestatementsopt
if-directive → #if
Platform condition Valid arguments elseif-directive → #elseif
else-directive → #else
os() macOS, iOS, watchOS, tvOS, Linux endif-directive → #endif
compilation-condition → platform-condition
arch() i386, x86_64, arm, arm64
compilation-condition → identifier
compilation-condition → boolean-literal
swift() >= followed by a version number
compilation-condition → (compilation-condition)
compilation-condition → !compilation-condition
The version number for the swift() platform condition consists of a major and minor
compilation-condition → compilation-condition&&compilation-condition
number, separated by a dot (.). There must not be whitespace between >= and the
compilation-condition → compilation-condition||compilation-condition
version number.
platform-condition → os(operating-system)
NOTE
platform-condition → arch(architecture)
platform-condition → swift(>=swift-version)
The arch(arm) platform condition does not return true for ARM 64 devices. The operating-system → macOS iOS watchOS tvOS
arch(i386) platform condition returns true when code is compiled for the 32–bit iOS
architecture → i386 x86_64 arm arm64
simulator.
swift-version → decimal-digits.decimal-digits
You can combine compilation conditions using the logical operators &&, ||, and ! and
use parentheses for grouping. Line Control Statement
A line control statement is used to specify a line number and filename that can be
Similar to an if statement, you can add multiple conditional branches to test for different from the line number and filename of the source code being compiled. Use a
different compilation conditions. You can add any number of additional branches using

216
line control statement to change the source code location used by Swift for diagnostic G R A M M A R O F A N AVA I L A B I L I T Y C O N D I T I O N
and debugging purposes. availability-condition → #available(availability-arguments)
availability-arguments → availability-argument availability-argument,availability-
A line control statement has the following forms: arguments
availability-argument → platform-nameplatform-version
#sourceLocation(file: filename, line: line number)
#sourceLocation()
availability-argument → *
platform-name → iOS iOSApplicationExtension
The first form of a line control statement changes the values of the #line and #file
platform-name → macOS macOSApplicationExtension
literal expressions, beginning with the line of code following the line control statement.
platform-name → watchOS
The line number changes the value of #line and is any integer literal greater than
platform-name → tvOS
zero. The filename changes the value of #file and is a string literal.
platform-version → decimal-digits
platform-version → decimal-digits.decimal-digits
The second form of a line control statement, #sourceLocation(), resets the source platform-version → decimal-digits.decimal-digits.decimal-digits
code location back to the default line numbering and filename.

G R A M M A R O F A L I N E C O N T R O L S TAT E M E N T

line-control-statement → #sourceLocation(file:file-name,line:line-number)
line-control-statement → #sourceLocation()
line-number → A decimal integer greater than zero
file-name → static-string-literal

Availability Condition
An availability condition is used as a condition of an if, while, and guard statement to
query the availability of APIs at runtime, based on specified platforms arguments.

An availability condition has the following form:

if #available(platform name version, ..., *) {

statements to execute if the APIs are available
} else {
fallback statements to execute if the APIs are unavailable
}
You use an availability condition to execute a block of code, depending on whether the
APIs you want to use are available at runtime. The compiler uses the information from
the availability condition when it verifies that the APIs in that block of code are
available.

The availability condition takes a comma-separated list of platform names and

versions. Use iOS, macOS, watchOS, and tvOS for the platform names, and include the
corresponding version numbers. The * argument is required and specifies that on any
other platform, the body of the code block guarded by the availability condition
executes on the minimum deployment target specified by your target.

Unlike Boolean conditions, you can’t combine availability conditions using logical
operators such as && and ||.

217
 Section 6 behavior by marking the declaration with an access-level modifier, as described in
Access Control Levels.

Declarations G R A M M A R O F A T O P - L E V E L D E C L A R AT I O N

top-level-declaration → statementsopt

Code Blocks
Declarations
A code block is used by a variety of declarations and control structures to group
statements together. It has the following form:
A declaration introduces a new name or construct into your program. For example,
you use declarations to introduce functions and methods, variables and constants,
{
and to define new, named enumeration, structure, class, and protocol types. You can statements
also use a declaration to extend the behavior of an existing named type and to import }
symbols into your program that are declared elsewhere. The statements inside a code block include declarations, expressions, and other kinds
of statements and are executed in order of their appearance in source code.
In Swift, most declarations are also definitions in the sense that they are implemented
or initialized at the same time they are declared. That said, because protocols don’t GRAMMAR OF A CODE BLOCK
implement their members, most protocol members are declarations only. For
code-block → {statementsopt}
convenience and because the distinction isn’t that important in Swift, the term
declaration covers both declarations and definitions.
Import Declaration
G R A M M A R O F A D E C L A R AT I O N An import declaration lets you access symbols that are declared outside the current
file. The basic form imports the entire module; it consists of the import keyword
declaration → import-declaration
followed by a module name:
declaration → constant-declaration
declaration → variable-declaration
import module
declaration → typealias-declaration
declaration → function-declaration Providing more detail limits which symbols are imported—you can specify a specific
declaration → enum-declaration submodule or a specific declaration within a module or submodule. When this detailed
declaration → struct-declaration form is used, only the imported symbol (and not the module that declares it) is made
declaration → class-declaration available in the current scope.
declaration → protocol-declaration
import import kind module.symbol name
declaration → initializer-declaration import module.submodule
declaration → deinitializer-declaration
declaration → extension-declaration G R A M M A R O F A N I M P O R T D E C L A R AT I O N
declaration → subscript-declaration import-declaration → attributesoptimportimport-kindoptimport-path
declaration → operator-declaration import-kind → typealias struct class enum protocol var func
declaration → precedence-group-declaration import-path → import-path-identifier import-path-identifier.import-path
declarations → declarationdeclarationsopt import-path-identifier → identifier operator

Top-Level Code Constant Declaration

The top-level code in a Swift source file consists of zero or more statements, A constant declaration introduces a constant named value into your program.
declarations, and expressions. By default, variables, constants, and other named Constant declarations are declared using the let keyword and have the following
declarations that are declared at the top-level of a source file are accessible to code in form:
every source file that is part of the same module. You can override this default

218
 let constant name: type = expression Variable declarations have several forms that declare different kinds of named,
A constant declaration defines an immutable binding between the constant name and mutable values, including stored and computed variables and properties, stored
the value of the initializer expression; after the value of a constant is set, it cannot be variable and property observers, and static variable properties. The appropriate form
changed. That said, if a constant is initialized with a class object, the object itself can to use depends on the scope at which the variable is declared and the kind of variable
change, but the binding between the constant name and the object it refers to can’t. you intend to declare.

When a constant is declared at global scope, it must be initialized with a value. When NOTE
a constant declaration occurs in the context of a function or method, it can be You can also declare properties in the context of a protocol declaration, as described in
initialized later, as long as it is guaranteed to have a value set before the first time its Protocol Property Declaration.
value is read. When a constant declaration occurs in the context of a class or structure
declaration, it is considered a constant property. Constant declarations are not You can override a property in a subclass by marking the subclass’s property
computed properties and therefore do not have getters or setters. declaration with the override declaration modifier, as described in Overriding.

If the constant name of a constant declaration is a tuple pattern, the name of each Stored Variables and Stored Variable Properties
item in the tuple is bound to the corresponding value in the initializer expression. The following form declares a stored variable or stored variable property:

let (firstNumber, secondNumber) = (10, 42) var variable name: type = expression

In this example, firstNumber is a named constant for the value 10, and secondNumber You define this form of a variable declaration at global scope, the local scope of a
is a named constant for the value 42. Both constants can now be used independently: function, or in the context of a class or structure declaration. When a variable
declaration of this form is declared at global scope or the local scope of a function, it is
print("The first number is \(firstNumber).") referred to as a stored variable. When it is declared in the context of a class or
structure declaration, it is referred to as a stored variable property.
// Prints "The first number is 10."
print("The second number is \(secondNumber).")
The initializer expression can’t be present in a protocol declaration, but in all other
// Prints "The second number is 42." contexts, the initializer expression is optional. That said, if no initializer expression is
The type annotation (: type) is optional in a constant declaration when the type of the present, the variable declaration must include an explicit type annotation (: type).
constant name can be inferred, as described in Type Inference.
As with constant declarations, if the variable name is a tuple pattern, the name of each
To declare a constant type property, mark the declaration with the static declaration item in the tuple is bound to the corresponding value in the initializer expression.
modifier. Type properties are discussed in Type Properties.
As their names suggest, the value of a stored variable or a stored variable property is
For more information about constants and for guidance about when to use them, see stored in memory.
Constants and Variables and Stored Properties.
Computed Variables and Computed Properties
G R A M M A R O F A C O N S TA N T D E C L A R AT I O N The following form declares a computed variable or computed property:
constant-declaration → attributesoptdeclaration-modifiersoptletpattern-initializer-list
var variable name: type {
pattern-initializer-list → pattern-initializer pattern-initializer,pattern-initializer-list get {
pattern-initializer → patterninitializeropt statements
initializer → =expression }
set(setter name) {
statements
Variable Declaration }
}
A variable declaration introduces a variable named value into your program and is
declared using the var keyword. You define this form of a variable declaration at global scope, the local scope of a
function, or in the context of a class, structure, enumeration, or extension declaration.

219
When a variable declaration of this form is declared at global scope or the local scope The initializer expression is optional in the context of a class or structure declaration,
of a function, it is referred to as a computed variable. When it is declared in the but required elsewhere. The type annotation is optional when the type can be inferred
context of a class, structure, or extension declaration, it is referred to as a computed from the initializer expression.
property.
The willSet and didSet observers provide a way to observe (and to respond
The getter is used to read the value, and the setter is used to write the value. The appropriately) when the value of a variable or property is being set. The observers are
setter clause is optional, and when only a getter is needed, you can omit both clauses not called when the variable or property is first initialized. Instead, they are called only
and simply return the requested value directly, as described in Read-Only Computed when the value is set outside of an initialization context.
Properties. But if you provide a setter clause, you must also provide a getter clause.
A willSet observer is called just before the value of the variable or property is set.
The setter name and enclosing parentheses is optional. If you provide a setter name, The new value is passed to the willSet observer as a constant, and therefore it can’t
it is used as the name of the parameter to the setter. If you do not provide a setter be changed in the implementation of the willSet clause. The didSet observer is
name, the default parameter name to the setter is newValue, as described in called immediately after the new value is set. In contrast to the willSet observer, the
Shorthand Setter Declaration. old value of the variable or property is passed to the didSet observer in case you still
need access to it. That said, if you assign a value to a variable or property within its
Unlike stored named values and stored variable properties, the value of a computed own didSet observer clause, that new value that you assign will replace the one that
named value or a computed property is not stored in memory. was just set and passed to the willSet observer.

For more information and to see examples of computed properties, see Computed The setter name and enclosing parentheses in the willSet and didSet clauses are
Properties. optional. If you provide setter names, they are used as the parameter names to the
willSet and didSet observers. If you do not provide setter names, the default
Stored Variable Observers and Property Observers parameter name to the willSet observer is newValue and the default parameter name
You can also declare a stored variable or property with willSet and didSet observers. to the didSet observer is oldValue.
A stored variable or property declared with observers has the following form:
The didSet clause is optional when you provide a willSet clause. Likewise, the
var variable name: type = expression { willSet clause is optional when you provide a didSet clause.
willSet(setter name) {
statements For more information and to see an example of how to use property observers, see
}
didSet(setter name) { Property Observers.
statements
} Type Variable Properties
}
To declare a type variable property, mark the declaration with the static declaration
You define this form of a variable declaration at global scope, the local scope of a
modifier. Classes may mark type computed properties with the class declaration
function, or in the context of a class or structure declaration. When a variable
modifier instead to allow subclasses to override the superclass’s implementation. Type
declaration of this form is declared at global scope or the local scope of a function, the
properties are discussed in Type Properties.
observers are referred to as stored variable observers. When it is declared in the
context of a class or structure declaration, the observers are referred to as property NOTE
observers.
In a class declaration, the static keyword has the same effect as marking the
declaration with both the class and final declaration modifiers.
You can add property observers to any stored property. You can also add property
observers to any inherited property (whether stored or computed) by overriding the G R A M M A R O F A VA R I A B L E D E C L A R AT I O N
property within a subclass, as described in Overriding Property Observers.
variable-declaration → variable-declaration-headpattern-initializer-list
variable-declaration → variable-declaration-headvariable-nametype-
annotationcode-block

220
 variable-declaration → variable-declaration-headvariable-nametype- var dictionary2: Dictionary<String, Int> = [:]
annotationgetter-setter-block When a type alias is declared with generic parameters, the constraints on those
variable-declaration → variable-declaration-headvariable-nametype- parameters must match exactly the constraints on the existing type’s generic
annotationgetter-setter-keyword-block parameters. For example:
variable-declaration → variable-declaration-headvariable-nameinitializerwillSet-
didSet-block typealias DictionaryOfInts<Key: Hashable> = Dictionary<Key, Int>
variable-declaration → variable-declaration-headvariable-nametype-
Because the type alias and the existing type can be used interchangeably, the type
annotationinitializeroptwillSet-didSet-block
alias can’t introduce additional generic constraints.
variable-declaration-head → attributesoptdeclaration-modifiersoptvar
variable-name → identifier
Inside a protocol declaration, a type alias can give a shorter and more convenient
getter-setter-block → code-block
name to a type that is used frequently. For example:
getter-setter-block → {getter-clausesetter-clauseopt}
getter-setter-block → {setter-clausegetter-clause}
protocol Sequence {
getter-clause → attributesoptmutation-modifieroptgetcode-block
setter-clause → attributesoptmutation-modifieroptsetsetter-nameoptcode-block associatedtype Iterator: IteratorProtocol
setter-name → (identifier) typealias Element = Iterator.Element
getter-setter-keyword-block → {getter-keyword-clausesetter-keyword-clauseopt} }
getter-setter-keyword-block → {setter-keyword-clausegetter-keyword-clause}
getter-keyword-clause → attributesoptmutation-modifieroptget
func sum<T: Sequence>(_ sequence: T) -> Int where T.Element == Int
setter-keyword-clause → attributesoptmutation-modifieroptset {
willSet-didSet-block → {willSet-clausedidSet-clauseopt}
// ...
willSet-didSet-block → {didSet-clausewillSet-clauseopt}
willSet-clause → attributesoptwillSetsetter-nameoptcode-block }
didSet-clause → attributesoptdidSetsetter-nameoptcode-block Without this type alias, the sum function would have to refer to the associated type as
T.Iterator.Element instead of T.Element.
Type Alias Declaration
A type alias declaration introduces a named alias of an existing type into your See also Protocol Associated Type Declaration.
program. Type alias declarations are declared using the typealias keyword and have
G R A M M A R O F A T Y P E A L I A S D E C L A R AT I O N
the following form:
typealias-declaration → attributesoptaccess-level-modifieropttypealiastypealias-
typealias name = existing type namegeneric-parameter-clauseopttypealias-assignment
After a type alias is declared, the aliased name can be used instead of the existing typealias-name → identifier
type everywhere in your program. The existing type can be a named type or a typealias-assignment → =type
compound type. Type aliases do not create new types; they simply allow a name to
refer to an existing type. Function Declaration
A function declaration introduces a function or method into your program. A function
A type alias declaration can use generic parameters to give a name to an existing
declared in the context of class, structure, enumeration, or protocol is referred to as a
generic type. The type alias can provide concrete types for some or all of the generic
method. Function declarations are declared using the func keyword and have the
parameters of the existing type. For example:
following form:
typealias StringDictionary<Value> = Dictionary<String, Value>
func function name(parameters) -> return type {
statements
}
// The following dictionaries have the same type.
var dictionary1: StringDictionary<Int> = [:]
If the function has a return type of Void, the return type can be omitted as follows:

221
 func function name(parameters) { In-out parameters are passed as follows:
statements
}
1.When the function is called, the value of the argument is copied.
The type of each parameter must be included—it can’t be inferred. If you write inout
in front of a parameter’s type, the parameter can be modified inside the scope of the 2.In the body of the function, the copy is modified.
function. In-out parameters are discussed in detail in In-Out Parameters, below. 3.When the function returns, the copy’s value is assigned to the original
argument.
Functions can return multiple values using a tuple type as the return type of the
function. This behavior is known as copy-in copy-out or call by value result. For example, when
a computed property or a property with observers is passed as an in-out parameter, its
A function definition can appear inside another function declaration. This kind of getter is called as part of the function call and its setter is called as part of the function
function is known as a nested function. For a discussion of nested functions, see return.
Nested Functions.
As an optimization, when the argument is a value stored at a physical address in
Parameter Names memory, the same memory location is used both inside and outside the function body.
The optimized behavior is known as call by reference; it satisfies all of the
Function parameters are a comma separated list where each parameter has one of
requirements of the copy-in copy-out model while removing the overhead of copying.
several forms. The order of arguments in a function call must match the order of
Write your code using the model given by copy-in copy-out, without depending on the
parameters in the function’s declaration. The simplest entry in a parameter list has the
call-by-reference optimization, so that it behaves correctly with or without the
following form:
optimization.
parameter name: parameter type
Do not access the value that was passed as an in-out argument, even if the original
A parameter has a name, which is used within the function body, as well as an
argument is available in the current scope. When the function returns, your changes to
argument label, which is used when calling the function or method. By default,
the original are overwritten with the value of the copy. Do not depend on the
parameter names are also used as argument labels. For example:
implementation of the call-by-reference optimization to try to keep the changes from
being overwritten.
func f(x: Int, y: Int) -> Int { return x + y }
f(x: 1, y: 2) // both x and y are labeled
You can’t pass the same argument to multiple in-out parameters because the order in
You can override the default behavior for argument labels with one of the following which the copies are written back is not well defined, which means the final value of
forms: the original would also not be well defined. For example:

argument label parameter name: parameter type var x = 10

_ parameter name: parameter type
func f(a: inout Int, b: inout Int) {
A name before the parameter name gives the parameter an explicit argument label,
a += 1
which can be different from the parameter name. The corresponding argument must
use the given argument label in function or method calls. b += 10
}
An underscore (_) before a parameter name suppresses the argument label. The f(a: &x, b: &x) // Invalid, in-out arguments alias each other
corresponding argument must have no label in function or method calls.
A closure or nested function that captures an in-out parameter must be nonescaping.
If you need to capture an in-out parameter without mutating it or to observe changes
func repeatGreeting(_ greeting: String, count n: Int) { /* Greet n
times */ } made by other code, use a capture list to explicitly capture the parameter immutably.
repeatGreeting("Hello, world!", count: 2) // count is labeled,
func someFunction(a: inout Int) -> () -> Int {
greeting is not
return { [a] in return a + 1 }
In-Out Parameters }

222
If you need to capture and mutate an in-out parameter, use an explicit local copy, such Methods on an enumeration or a structure that modify self must be marked with the
as in multithreaded code that ensures all mutation has finished before the function mutating declaration modifier.
returns.
Methods that override a superclass method must be marked with the override
func multithreadedFunction(queue: DispatchQueue, x: inout Int) { declaration modifier. It’s a compile-time error to override a method without the
// Make a local copy and manually copy it back. override modifier or to use the override modifier on a method that doesn’t override a
var localX = x superclass method.
defer { x = localX }
Methods associated with a type rather than an instance of a type must be marked with
the static declaration modifier for enumerations and structures or the class
// Operate on localX asynchronously, then wait before declaration modifier for classes.
returning.
queue.async { someMutatingOperation(&localX) } Throwing Functions and Methods
queue.sync {} Functions and methods that can throw an error must be marked with the throws
} keyword. These functions and methods are known as throwing functions and throwing
For more discussion and examples of in-out parameters, see In-Out Parameters. methods. They have the following form:

func function name(parameters) throws -> return type {

Special Kinds of Parameters statements
Parameters can be ignored, take a variable number of values, and provide default }
values using the following forms: Calls to a throwing function or method must be wrapped in a try or try! expression
(that is, in the scope of a try or try! operator).
_ : parameter type
parameter name: parameter type...
parameter name: parameter type = default argument value
The throws keyword is part of a function’s type, and nonthrowing functions are
subtypes of throwing functions. As a result, you can use a nonthrowing function in the
An underscore (_) parameter is explicitly ignored and can’t be accessed within the
same places as a throwing one.
body of the function.
You can’t overload a function based only on whether the function can throw an error.
A parameter with a base type name followed immediately by three dots (...) is
That said, you can overload a function based on whether a function parameter can
understood as a variadic parameter. A function can have at most one variadic
throw an error.
parameter. A variadic parameter is treated as an array that contains elements of the
base type name. For instance, the variadic parameter Int... is treated as [Int]. For
A throwing method can’t override a nonthrowing method, and a throwing method can’t
an example that uses a variadic parameter, see Variadic Parameters.
satisfy a protocol requirement for a nonthrowing method. That said, a nonthrowing
method can override a throwing method, and a nonthrowing method can satisfy a
A parameter with an equals sign (=) and an expression after its type is understood to
protocol requirement for a throwing method.
have a default value of the given expression. The given expression is evaluated when
the function is called. If the parameter is omitted when calling the function, the default
Rethrowing Functions and Methods
value is used instead.
A function or method can be declared with the rethrows keyword to indicate that it
func f(x: Int = 42) -> Int { return x } throws an error only if one of its function parameters throws an error. These functions
f() // Valid, uses default value
and methods are known as rethrowing functions and rethrowing methods. Rethrowing
functions and methods must have at least one throwing function parameter.
f(x: 7) // Valid, uses the value provided
f(7) // Invalid, missing argument label func someFunction(callback: () throws -> Void) rethrows {
try callback()
Special Kinds of Methods
}

223
A rethrowing function or method can contain a throw statement only inside a catch function-declaration → function-headfunction-namegeneric-parameter-
clause. This lets you call the throwing function inside a do-catch block and handle clauseoptfunction-signaturegeneric-where-clauseoptfunction-bodyopt
errors in the catch clause by throwing a different error. In addition, the catch clause function-head → attributesoptdeclaration-modifiersoptfunc
must handle only errors thrown by one of the rethrowing function’s throwing function-name → identifier operator
parameters. For example, the following is invalid because the catch clause would function-signature → parameter-clausethrowsoptfunction-resultopt
handle the error thrown by alwaysThrows(). function-signature → parameter-clauserethrowsfunction-resultopt
function-result → ->attributesopttype
func alwaysThrows() throws { function-body → code-block
throw SomeError.error parameter-clause → () (parameter-list)
} parameter-list → parameter parameter,parameter-list
parameter → external-parameter-nameoptlocal-parameter-nametype-
func someFunction(callback: () throws -> Void) rethrows {
annotationdefault-argument-clauseopt
do { parameter → external-parameter-nameoptlocal-parameter-nametype-annotation
try callback() parameter → external-parameter-nameoptlocal-parameter-nametype-annotation...
try alwaysThrows() // Invalid, alwaysThrows() isn't a external-parameter-name → identifier
throwing parameter local-parameter-name → identifier
} catch { default-argument-clause → =expression
throw AnotherError.error
}
Enumeration Declaration
}
An enumeration declaration introduces a named enumeration type into your program.

Enumeration declarations have two basic forms and are declared using the enum
A throwing method can’t override a rethrowing method, and a throwing method can’t keyword. The body of an enumeration declared using either form contains zero or
satisfy a protocol requirement for a rethrowing method. That said, a rethrowing more values—called enumeration cases—and any number of declarations, including
method can override a throwing method, and a rethrowing method can satisfy a computed properties, instance methods, type methods, initializers, type aliases, and
protocol requirement for a throwing method. even other enumeration, structure, and class declarations. Enumeration declarations
can’t contain deinitializer or protocol declarations.
Functions that Never Return
Swift defines a Never type, which indicates that a function or method doesn’t return to Enumeration types can adopt any number of protocols, but can’t inherit from classes,
its caller. Functions and methods with the Never return type are called nonreturning. structures, or other enumerations.
Nonreturning functions and methods either cause an irrecoverable error or begin a
sequence of work that continues indefinitely. This means that code that would Unlike classes and structures, enumeration types do not have an implicitly provided
otherwise run immediately after the call is never executed. Throwing and rethrowing default initializer; all initializers must be declared explicitly. Initializers can delegate to
functions can transfer program control to an appropriate catch block, even when they other initializers in the enumeration, but the initialization process is complete only after
are nonreturning. an initializer assigns one of the enumeration cases to self.

A nonreturning function or method can be called to conclude the else clause of a Like structures but unlike classes, enumerations are value types; instances of an
guard statement, as discussed in Guard Statement. enumeration are copied when assigned to variables or constants, or when passed as
arguments to a function call. For information about value types, see Structures and
You can override a nonreturning method, but the new method must preserve its return Enumerations Are Value Types.
type and nonreturning behavior.
You can extend the behavior of an enumeration type with an extension declaration, as
G R A M M A R O F A F U N C T I O N D E C L A R AT I O N discussed in Extension Declaration.

Enumerations with Cases of Any Type

224
The following form declares an enumeration type that contains enumeration cases of case empty
any type: indirect case node(value: T, left: Tree, right: Tree)
}
enum enumeration name: adopted protocols {
case enumeration case 1 To enable indirection for all the cases of an enumeration, mark the entire enumeration
case enumeration case 2(associated value types) with the indirect modifier—this is convenient when the enumeration contains many
}
cases that would each need to be marked with the indirect modifier.
Enumerations declared in this form are sometimes called discriminated unions in other
programming languages. An enumeration case that’s marked with the indirect modifier must have an
associated value. An enumeration that is marked with the indirect modifier can
In this form, each case block consists of the case keyword followed by one or more contain a mixture of cases that have associated values and cases those that don’t.
enumeration cases, separated by commas. The name of each case must be unique. That said, it can’t contain any cases that are also marked with the indirect modifier.
Each case can also specify that it stores values of a given type. These types are
specified in the associated value types tuple, immediately following the name of the Enumerations with Cases of a Raw-Value Type
case.
The following form declares an enumeration type that contains enumeration cases of
the same basic type:
Enumeration cases that store associated values can be used as functions that create
instances of the enumeration with the specified associated values. And just like enum enumeration name: raw-value type, adopted protocols {
functions, you can get a reference to an enumeration case and apply it later in your case enumeration case 1 = raw value 1
code. case enumeration case 2 = raw value 2
}
enum Number { In this form, each case block consists of the case keyword, followed by one or more
case integer(Int) enumeration cases, separated by commas. Unlike the cases in the first form, each
case has an underlying value, called a raw value, of the same basic type. The type of
case real(Double)
these values is specified in the raw-value type and must represent an integer, floating-
} point number, string, or single character. In particular, the raw-value type must
let f = Number.integer conform to the Equatable protocol and one of the following protocols:
// f is a function of type (Int) -> Number ExpressibleByIntegerLiteral for integer literals, ExpressibleByFloatLiteral for
floating-point literals, ExpressibleByStringLiteral for string literals that contain any
number of characters, and ExpressibleByUnicodeScalarLiteral or
// Apply f to create an array of Number instances with integer
values ExpressibleByExtendedGraphemeClusterLiteral for string literals that contain only a
single character. Each case must have a unique name and be assigned a unique raw
let evenInts: [Number] = [0, 2, 4, 6].map(f)
value.
For more information and to see examples of cases with associated value types, see
Associated Values. If the raw-value type is specified as Int and you don’t assign a value to the cases
explicitly, they are implicitly assigned the values 0, 1, 2, and so on. Each unassigned
Enumerations with Indirection case of type Int is implicitly assigned a raw value that is automatically incremented
Enumerations can have a recursive structure, that is, they can have cases with from the raw value of the previous case.
associated values that are instances of the enumeration type itself. However,
instances of enumeration types have value semantics, which means they have a fixed enum ExampleEnum: Int {
layout in memory. To support recursion, the compiler must insert a layer of indirection. case a, b, c = 5, d
}
To enable indirection for a particular enumeration case, mark it with the indirect
In the above example, the raw value of ExampleEnum.a is 0 and the value of
declaration modifier.
ExampleEnum.b is 1. And because the value of ExampleEnum.c is explicitly set to 5, the

enum Tree<T> {
value of ExampleEnum.d is automatically incremented from 5 and is therefore 6.

225
If the raw-value type is specified as String and you don’t assign values to the cases union-style-enum-case → enum-case-nametuple-typeopt
explicitly, each unassigned case is implicitly assigned a string with the same text as enum-name → identifier
the name of that case. enum-case-name → identifier
raw-value-style-enum → enumenum-namegeneric-parameter-clauseopttype-
enum GamePlayMode: String { inheritance-clausegeneric-where-clauseopt{raw-value-style-enum-members}
case cooperative, individual, competitive raw-value-style-enum-members → raw-value-style-enum-memberraw-value-style-
} enum-membersopt
raw-value-style-enum-member → declaration raw-value-style-enum-case-clause
In the above example, the raw value of GamePlayMode.cooperative is "cooperative", compiler-control-statement
the raw value of GamePlayMode.individual is "individual",. and the raw value of raw-value-style-enum-case-clause → attributesoptcaseraw-value-style-enum-case-
GamePlayMode.competitive is "competitive". list
raw-value-style-enum-case-list → raw-value-style-enum-case raw-value-style-
Enumerations that have cases of a raw-value type implicitly conform to the enum-case,raw-value-style-enum-case-list
RawRepresentable protocol, defined in the Swift standard library. As a result, they raw-value-style-enum-case → enum-case-nameraw-value-assignmentopt
have a rawValue property and a failable initializer with the signature init?(rawValue: raw-value-assignment → =raw-value-literal
RawValue). You can use the rawValue property to access the raw value of an raw-value-literal → numeric-literal static-string-literal boolean-literal
enumeration case, as in ExampleEnum.B.rawValue. You can also use a raw value to
find a corresponding case, if there is one, by calling the enumeration’s failable
Structure Declaration
initializer, as in ExampleEnum(rawValue: 5), which returns an optional case. For more
information and to see examples of cases with raw-value types, see Raw Values. A structure declaration introduces a named structure type into your program. Structure
declarations are declared using the struct keyword and have the following form:
Accessing Enumeration Cases
struct structure name: adopted protocols {
To reference the case of an enumeration type, use dot (.) syntax, as in declarations
EnumerationType.enumerationCase. When the enumeration type can be inferred from }
context, you can omit it (the dot is still required), as described in Enumeration Syntax The body of a structure contains zero or more declarations. These declarations can
and Implicit Member Expression. include both stored and computed properties, type properties, instance methods, type
methods, initializers, subscripts, type aliases, and even other structure, class, and
To check the values of enumeration cases, use a switch statement, as shown in enumeration declarations. Structure declarations can’t contain deinitializer or protocol
Matching Enumeration Values with a Switch Statement. The enumeration type is declarations. For a discussion and several examples of structures that include various
pattern-matched against the enumeration case patterns in the case blocks of the kinds of declarations, see Classes and Structures.
switch statement, as described in Enumeration Case Pattern.
Structure types can adopt any number of protocols, but can’t inherit from classes,
G R A M M A R O F A N E N U M E R AT I O N D E C L A R AT I O N enumerations, or other structures.
enum-declaration → attributesoptaccess-level-modifieroptunion-style-enum
enum-declaration → attributesoptaccess-level-modifieroptraw-value-style-enum There are three ways create an instance of a previously declared structure:
union-style-enum → indirectoptenumenum-namegeneric-parameter-clauseopttype-
inheritance-clauseoptgeneric-where-clauseopt{union-style-enum-membersopt} Call one of the initializers declared within the structure, as described in
union-style-enum-members → union-style-enum-memberunion-style-enum- Initializers.
membersopt If no initializers are declared, call the structure’s memberwise initializer, as
union-style-enum-member → declaration union-style-enum-case-clause compiler- described in Memberwise Initializers for Structure Types.
control-statement
union-style-enum-case-clause → attributesoptindirectoptcaseunion-style-enum- If no initializers are declared, and all properties of the structure declaration
case-list were given initial values, call the structure’s default initializer, as described
union-style-enum-case-list → union-style-enum-case union-style-enum- in Default Initializers.
case,union-style-enum-case-list

226
The process of initializing a structure’s declared properties is described in properties and it must do so before calling any of its superclass’s designated
Initialization. initializers.

Properties of a structure instance can be accessed using dot (.) syntax, as described A class can override properties, methods, subscripts, and initializers of its superclass.
in Accessing Properties. Overridden properties, methods, subscripts, and designated initializers must be
marked with the override declaration modifier.
Structures are value types; instances of a structure are copied when assigned to
variables or constants, or when passed as arguments to a function call. For To require that subclasses implement a superclass’s initializer, mark the superclass’s
information about value types, see Structures and Enumerations Are Value Types. initializer with the required declaration modifier. The subclass’s implementation of that
initializer must also be marked with the required declaration modifier.
You can extend the behavior of a structure type with an extension declaration, as
discussed in Extension Declaration. Although properties and methods declared in the superclass are inherited by the
current class, designated initializers declared in the superclass are only inherited
G R A M M A R O F A S T R U C T U R E D E C L A R AT I O N when the subclass meets the conditions described in Automatic Initializer Inheritance.
Swift classes do not inherit from a universal base class.
struct-declaration → attributesoptaccess-level-modifieroptstructstruct-namegeneric-
parameter-clauseopttype-inheritance-clauseoptgeneric-where-clauseoptstruct-body
struct-name → identifier There are two ways create an instance of a previously declared class:
struct-body → {struct-membersopt}
struct-members → struct-memberstruct-membersopt Call one of the initializers declared within the class, as described in
struct-member → declaration compiler-control-statement Initializers.
If no initializers are declared, and all properties of the class declaration
Class Declaration were given initial values, call the class’s default initializer, as described in
A class declaration introduces a named class type into your program. Class Default Initializers.
declarations are declared using the class keyword and have the following form: Access properties of a class instance with dot (.) syntax, as described in Accessing
Properties.
class class name: superclass, adopted protocols {
declarations
} Classes are reference types; instances of a class are referred to, rather than copied,
when assigned to variables or constants, or when passed as arguments to a function
The body of a class contains zero or more declarations. These declarations can
call. For information about reference types, see Structures and Enumerations Are
include both stored and computed properties, instance methods, type methods,
Value Types.
initializers, a single deinitializer, subscripts, type aliases, and even other class,
structure, and enumeration declarations. Class declarations can’t contain protocol
declarations. For a discussion and several examples of classes that include various You can extend the behavior of a class type with an extension declaration, as
kinds of declarations, see Classes and Structures. discussed in Extension Declaration.

G R A M M A R O F A C L A S S D E C L A R AT I O N
A class type can inherit from only one parent class, its superclass, but can adopt any
number of protocols. The superclass appears first after the class name and colon, class-declaration → attributesoptaccess-level-modifieroptfinaloptclassclass-
followed by any adopted protocols. Generic classes can inherit from other generic and namegeneric-parameter-clauseopttype-inheritance-clauseoptgeneric-where-
nongeneric classes, but a nongeneric class can inherit only from other nongeneric clauseoptclass-body
classes. When you write the name of a generic superclass class after the colon, you class-declaration → attributesoptfinalaccess-level-modifieroptclassclass-
must include the full name of that generic class, including its generic parameter namegeneric-parameter-clauseopttype-inheritance-clauseoptgeneric-where-
clause. clauseoptclass-body
class-name → identifier
As discussed in Initializer Declaration, classes can have designated and convenience class-body → {class-membersopt}
initializers. The designated initializer of a class must initialize all of the class’s declared class-members → class-memberclass-membersopt

227
 class-member → declaration compiler-control-statement To restrict the adoption of a protocol to class types only, mark the protocol with the
class requirement by writing the class keyword as the first item in the inherited
Protocol Declaration protocols list after the colon. For example, the following protocol can be adopted only
by class types:
A protocol declaration introduces a named protocol type into your program. Protocol
declarations are declared at global scope using the protocol keyword and have the
protocol SomeProtocol: class {
following form:
/* Protocol members go here */
protocol protocol name: inherited protocols { }
protocol member declarations
} Any protocol that inherits from a protocol that’s marked with the class requirement
The body of a protocol contains zero or more protocol member declarations, which can likewise be adopted only by class types.
describe the conformance requirements that any type adopting the protocol must
NOTE
fulfill. In particular, a protocol can declare that conforming types must implement
certain properties, methods, initializers, and subscripts. Protocols can also declare If a protocol is marked with the objc attribute, the class requirement is implicitly
special kinds of type aliases, called associated types, that can specify relationships applied to that protocol; there’s no need to mark the protocol with the class
requirement explicitly.
among the various declarations of the protocol. Protocol declarations can’t contain
class, structure, enumeration, or other protocol declarations. The protocol member Protocols are named types, and thus they can appear in all the same places in your
declarations are discussed in detail below. code as other named types, as discussed in Protocols as Types. However, you can’t
construct an instance of a protocol, because protocols do not actually provide the
Protocol types can inherit from any number of other protocols. When a protocol type implementations for the requirements they specify.
inherits from other protocols, the set of requirements from those other protocols are
aggregated, and any type that inherits from the current protocol must conform to all You can use protocols to declare which methods a delegate of a class or structure
those requirements. For an example of how to use protocol inheritance, see Protocol should implement, as described in Delegation.
Inheritance.
G R A M M A R O F A P R O T O C O L D E C L A R AT I O N
NOTE
protocol-declaration → attributesoptaccess-level-modifieroptprotocolprotocol-
You can also aggregate the conformance requirements of multiple protocols using nametype-inheritance-clauseoptprotocol-body
protocol composition types, as described in Protocol Composition Type and Protocol
protocol-name → identifier
Composition.
protocol-body → {protocol-membersopt}
You can add protocol conformance to a previously declared type by adopting the protocol-members → protocol-memberprotocol-membersopt
protocol in an extension declaration of that type. In the extension, you must implement protocol-member → protocol-member-declaration compiler-control-statement
all of the adopted protocol’s requirements. If the type already implements all of the protocol-member-declaration → protocol-property-declaration
requirements, you can leave the body of the extension declaration empty. protocol-member-declaration → protocol-method-declaration
protocol-member-declaration → protocol-initializer-declaration
By default, types that conform to a protocol must implement all properties, methods, protocol-member-declaration → protocol-subscript-declaration
and subscripts declared in the protocol. That said, you can mark these protocol protocol-member-declaration → protocol-associated-type-declaration
member declarations with the optional declaration modifier to specify that their protocol-member-declaration → typealias-declaration
implementation by a conforming type is optional. The optional modifier can be
applied only to members that are marked with the objc attribute, and only to members Protocol Property Declaration
of protocols that are marked with the objc attribute. As a result, only class types can Protocols declare that conforming types must implement a property by including a
adopt and conform to a protocol that contains optional member requirements. For protocol property declaration in the body of the protocol declaration. Protocol property
more information about how to use the optional declaration modifier and for guidance declarations have a special form of a variable declaration:
about how to access optional protocol members—for example, when you’re not sure
whether a conforming type implements them—see Optional Protocol Requirements. var property name: type { get set }

228
As with other protocol member declarations, these property declarations declare only initializer declarations have the same form as initializer declarations, except they don’t
the getter and setter requirements for types that conform to the protocol. As a result, include the initializer’s body.
you don’t implement the getter or setter directly in the protocol in which it is declared.
A conforming type can satisfy a nonfailable protocol initializer requirement by
The getter and setter requirements can be satisfied by a conforming type in a variety implementing a nonfailable initializer or an init! failable initializer. A conforming type
of ways. If a property declaration includes both the get and set keywords, a can satisfy a failable protocol initializer requirement by implementing any kind of
conforming type can implement it with a stored variable property or a computed initializer.
property that is both readable and writeable (that is, one that implements both a getter
and a setter). However, that property declaration can’t be implemented as a constant When a class implements an initializer to satisfy a protocol’s initializer requirement,
property or a read-only computed property. If a property declaration includes only the the initializer must be marked with the required declaration modifier if the class is not
get keyword, it can be implemented as any kind of property. For examples of already marked with the final declaration modifier.
conforming types that implement the property requirements of a protocol, see Property
Requirements. See also Initializer Declaration.

See also Variable Declaration. G R A M M A R O F A P R O T O C O L I N I T I A L I Z E R D E C L A R AT I O N

G R A M M A R O F A P R O T O C O L P R O P E R T Y D E C L A R AT I O N
protocol-initializer-declaration → initializer-headgeneric-parameter-
clauseoptparameter-clausethrowsoptgeneric-where-clauseopt
protocol-property-declaration → variable-declaration-headvariable-nametype- protocol-initializer-declaration → initializer-headgeneric-parameter-
annotationgetter-setter-keyword-block clauseoptparameter-clauserethrowsgeneric-where-clauseopt

Protocol Method Declaration Protocol Subscript Declaration

Protocols declare that conforming types must implement a method by including a Protocols declare that conforming types must implement a subscript by including a
protocol method declaration in the body of the protocol declaration. Protocol method protocol subscript declaration in the body of the protocol declaration. Protocol
declarations have the same form as function declarations, with two exceptions: They subscript declarations have a special form of a subscript declaration:
don’t include a function body, and you can’t provide any default parameter values as
part of the function declaration. For examples of conforming types that implement the subscript (parameters) -> return type { get set }
method requirements of a protocol, see Method Requirements. Subscript declarations only declare the minimum getter and setter implementation
requirements for types that conform to the protocol. If the subscript declaration
To declare a class or static method requirement in a protocol declaration, mark the includes both the get and set keywords, a conforming type must implement both a
method declaration with the static declaration modifier. Classes that implement this getter and a setter clause. If the subscript declaration includes only the get keyword, a
method declare the method with the class modifier. Structures that implement it must conforming type must implement at least a getter clause and optionally can implement
declare the method with the static declaration modifier instead. If you’re a setter clause.
implementing the method in an extension, use the class modifier if you’re extending a
class and the static modifier if you’re extending a structure. See also Subscript Declaration.

See also Function Declaration. G R A M M A R O F A P R O T O C O L S U B S C R I P T D E C L A R AT I O N

G R A M M A R O F A P R O T O C O L M E T H O D D E C L A R AT I O N
protocol-subscript-declaration → subscript-headsubscript-resultgetter-setter-
keyword-block
protocol-method-declaration → function-headfunction-namegeneric-parameter-
clauseoptfunction-signaturegeneric-where-clauseopt Protocol Associated Type Declaration
Protocols declare associated types using the associatedtype keyword. An associated
Protocol Initializer Declaration type provides an alias for a type that is used as part of a protocol’s declaration.
Protocols declare that conforming types must implement an initializer by including a Associated types are similar to type parameters in generic parameter clauses, but
protocol initializer declaration in the body of the protocol declaration. Protocol they’re associated with Self in the protocol in which they’re declared. In that context,

229
Self refers to the eventual type that conforms to the protocol. For more information Convenience initializers can delegate the initialization process to another convenience
and examples, see Associated Types. initializer or to one of the class’s designated initializers. That said, the initialization
processes must end with a call to a designated initializer that ultimately initializes the
See also Type Alias Declaration. class’s properties. Convenience initializers can’t call a superclass’s initializers.

G R A M M A R O F A P R O T O C O L A S S O C I AT E D T Y P E D E C L A R AT I O N You can mark designated and convenience initializers with the required declaration
modifier to require that every subclass implement the initializer. A subclass’s
protocol-associated-type-declaration → attributesoptaccess-level-
implementation of that initializer must also be marked with the required declaration
modifieroptassociatedtypetypealias-nametype-inheritance-clauseopttypealias-
modifier.
assignmentopt

By default, initializers declared in a superclass are not inherited by subclasses. That

Initializer Declaration said, if a subclass initializes all of its stored properties with default values and doesn’t
An initializer declaration introduces an initializer for a class, structure, or enumeration define any initializers of its own, it inherits all of the superclass’s initializers. If the
into your program. Initializer declarations are declared using the init keyword and subclass overrides all of the superclass’s designated initializers, it inherits the
have two basic forms. superclass’s convenience initializers.

Structure, enumeration, and class types can have any number of initializers, but the As with methods, properties, and subscripts, you need to mark overridden designated
rules and associated behavior for class initializers are different. Unlike structures and initializers with the override declaration modifier.
enumerations, classes have two kinds of initializers: designated initializers and
convenience initializers, as described in Initialization. NOTE

If you mark an initializer with the required declaration modifier, you don’t also mark the
The following form declares initializers for structures, enumerations, and designated initializer with the override modifier when you override the required initializer in a
initializers of classes: subclass.

init(parameters) { Just like functions and methods, initializers can throw or rethrow errors. And just like
statements functions and methods, you use the throws or rethrows keyword after an initializer’s
} parameters to indicate the appropriate behavior.
A designated initializer of a class initializes all of the class’s properties directly. It can’t
call any other initializers of the same class, and if the class has a superclass, it must To see examples of initializers in various type declarations, see Initialization.
call one of the superclass’s designated initializers. If the class inherits any properties
from its superclass, one of the superclass’s designated initializers must be called Failable Initializers
before any of these properties can be set or modified in the current class. A failable initializer is a type of initializer that produces an optional instance or an
implicitly unwrapped optional instance of the type the initializer is declared on. As a
Designated initializers can be declared in the context of a class declaration only and result, a failable initializer can return nil to indicate that initialization failed.
therefore can’t be added to a class using an extension declaration.
To declare a failable initializer that produces an optional instance, append a question
Initializers in structures and enumerations can call other declared initializers to mark to the init keyword in the initializer declaration (init?). To declare a failable
delegate part or all of the initialization process. initializer that produces an implicitly unwrapped optional instance, append an
exclamation mark instead (init!). The example below shows an init? failable
To declare convenience initializers for a class, mark the initializer declaration with the initializer that produces an optional instance of a structure.
convenience declaration modifier.
struct SomeStruct {
convenience init(parameters) {
statements let property: String
} // produces an optional instance of 'SomeStruct'
init?(input: String) {

230
 if input.isEmpty { initializer-head → attributesoptdeclaration-modifiersoptinit
// discard 'self' and return 'nil' initializer-head → attributesoptdeclaration-modifiersoptinit?
return nil initializer-head → attributesoptdeclaration-modifiersoptinit!
initializer-body → code-block
}
property = input
Deinitializer Declaration
}
A deinitializer declaration declares a deinitializer for a class type. Deinitializers take no
}
parameters and have the following form:
You call an init? failable initializer in the same way that you call a nonfailable
initializer, except that you must deal with the optionality of the result. deinit {
statements
}
if let actualInstance = SomeStruct(input: "Hello") {
A deinitializer is called automatically when there are no longer any references to a
// do something with the instance of 'SomeStruct'
class object, just before the class object is deallocated. A deinitializer can be declared
} else { only in the body of a class declaration—but not in an extension of a class—and each
// initialization of 'SomeStruct' failed and the initializer class can have at most one.
returned 'nil'
} A subclass inherits its superclass’s deinitializer, which is implicitly called just before
A failable initializer can return nil at any point in the implementation of the initializer’s the subclass object is deallocated. The subclass object is not deallocated until all
body. deinitializers in its inheritance chain have finished executing.

A failable initializer can delegate to any kind of initializer. A nonfailable initializer can Deinitializers are not called directly.
delegate to another nonfailable initializer or to an init! failable initializer. A nonfailable
initializer can delegate to an init? failable initializer by force-unwrapping the result of For an example of how to use a deinitializer in a class declaration, see Deinitialization.
the superclass’s initializer—for example, by writing super.init()!.
G R A M M A R O F A D E I N I T I A L I Z E R D E C L A R AT I O N

Initialization failure propagates through initializer delegation. Specifically, if a failable deinitializer-declaration → attributesoptdeinitcode-block
initializer delegates to an initializer that fails and returns nil, then the initializer that
delegated also fails and implicitly returns nil. If a nonfailable initializer delegates to an Extension Declaration
init! failable initializer that fails and returns nil, then a runtime error is raised (as if
you used the ! operator to unwrap an optional that has a nil value). An extension declaration allows you to extend the behavior of existing class, structure,
and enumeration types. Extension declarations are declared using the extension
keyword and have the following form:
A failable designated initializer can be overridden in a subclass by any kind of
designated initializer. A nonfailable designated initializer can be overridden in a
extension type name: adopted protocols {
subclass by a nonfailable designated initializer only. declarations
}
For more information and to see examples of failable initializers, see Failable The body of an extension declaration contains zero or more declarations. These
Initializers. declarations can include computed properties, computed type properties, instance
methods, type methods, initializers, subscript declarations, and even class, structure,
G R A M M A R O F A N I N I T I A L I Z E R D E C L A R AT I O N and enumeration declarations. Extension declarations can’t contain deinitializer or
initializer-declaration → initializer-headgeneric-parameter-clauseoptparameter- protocol declarations, stored properties, property observers, or other extension
clausethrowsoptgeneric-where-clauseoptinitializer-body declarations. For a discussion and several examples of extensions that include
initializer-declaration → initializer-headgeneric-parameter-clauseoptparameter- various kinds of declarations, see Extensions.
clauserethrowsgeneric-where-clauseoptinitializer-body

231
Extension declarations can add protocol conformance to an existing class, structure, As with computed properties, subscript declarations support reading and writing the
and enumeration type in the adopted protocols. Extension declarations can’t add class value of the accessed elements. The getter is used to read the value, and the setter is
inheritance to an existing class, and therefore you can specify only a list of protocols used to write the value. The setter clause is optional, and when only a getter is
after the type name and colon. needed, you can omit both clauses and simply return the requested value directly.
That said, if you provide a setter clause, you must also provide a getter clause.
Properties, methods, and initializers of an existing type can’t be overridden in an
extension of that type. The setter name and enclosing parentheses are optional. If you provide a setter
name, it is used as the name of the parameter to the setter. If you do not provide a
Extension declarations can contain initializer declarations. That said, if the type you’re setter name, the default parameter name to the setter is value. The type of the setter
extending is defined in another module, an initializer declaration must delegate to an name must be the same as the return type.
initializer already defined in that module to ensure members of that type are properly
initialized. You can overload a subscript declaration in the type in which it is declared, as long as
the parameters or the return type differ from the one you’re overloading. You can also
G R A M M A R O F A N E X T E N S I O N D E C L A R AT I O N override a subscript declaration inherited from a superclass. When you do so, you
must mark the overridden subscript declaration with the override declaration modifier.
extension-declaration → attributesoptaccess-level-modifieroptextensiontype-
identifiertype-inheritance-clauseoptextension-body
extension-declaration → attributesoptaccess-level-modifieroptextensiontype- By default, the parameters used in subscripting don’t have argument labels, unlike
identifiergeneric-where-clauseextension-body functions, methods, and initializers. However, you can provide explicit argument labels
extension-body → {extension-membersopt} using the same syntax that functions, methods, and initializers use.
extension-members → extension-memberextension-membersopt
extension-member → declaration compiler-control-statement You can also declare subscripts in the context of a protocol declaration, as described
in Protocol Subscript Declaration.
Subscript Declaration
For more information about subscripting and to see examples of subscript
A subscript declaration allows you to add subscripting support for objects of a declarations, see Subscripts.
particular type and are typically used to provide a convenient syntax for accessing the
elements in a collection, list, or sequence. Subscript declarations are declared using G R A M M A R O F A S U B S C R I P T D E C L A R AT I O N
the subscript keyword and have the following form:
subscript-declaration → subscript-headsubscript-resultcode-block
subscript (parameters) -> return type { subscript-declaration → subscript-headsubscript-resultgetter-setter-block
get { subscript-declaration → subscript-headsubscript-resultgetter-setter-keyword-block
statements subscript-head → attributesoptdeclaration-modifiersoptsubscriptparameter-clause
}
set(setter name) { subscript-result → ->attributesopttype
statements
}
}
Operator Declaration
Subscript declarations can appear only in the context of a class, structure, An operator declaration introduces a new infix, prefix, or postfix operator into your
enumeration, extension, or protocol declaration. program and is declared using the operator keyword.

The parameters specify one or more indexes used to access elements of the You can declare operators of three different fixities: infix, prefix, and postfix. The fixity
corresponding type in a subscript expression (for example, the i in the expression of an operator specifies the relative position of an operator to its operands.
object[i]). Although the indexes used to access the elements can be of any type,
each parameter must include a type annotation to specify the type of each index. The There are three basic forms of an operator declaration, one for each fixity. The fixity of
return type specifies the type of the element being accessed. the operator is specified by marking the operator declaration with the infix, prefix,
or postfix declaration modifier before the operator keyword. In each form, the name
of the operator can contain only the operator characters defined in Operators.

232
The following form declares a new infix operator: Precedence Group Declaration
A precedence group declaration introduces a new grouping for infix operator
infix operator operator name: precedence group
precedence into your program. The precedence of an operator specifies how tightly
An infix operator is a binary operator that is written between its two operands, such as the operator binds to its operands, in the absence of grouping parentheses.
the familiar addition operator (+) in the expression 1 + 2.
A precedence group declaration has the following form:
Infix operators can optionally specify a precedence group. If you omit the precedence
group for an operator, Swift uses the default precedence group, DefaultPrecedence, precedencegroup precedence group name {
which specifies a precedence just higher than TernaryPrecedence. For more higherThan: lower group names
information, see Precedence Group Declaration. lowerThan: higher group names
associativity: associativity
assignment: assignment
The following form declares a new prefix operator: }
The lower group names and higher group names lists specify the new precedence
prefix operator operator name
group’s relation to existing precedence groups. The lowerThan precedence group
A prefix operator is a unary operator that is written immediately before its operand, attribute may only be used to refer to precedence groups declared outside of the
such as the prefix logical NOT operator (!) in the expression !a. current module. When two operators compete with each other for their operands, such
as in the expression 2 + 3 * 5, the operator with the higher relative precedence binds
Prefix operators declarations don’t specify a precedence level. Prefix operators are more tightly to its operands.
nonassociative.
NOTE
The following form declares a new postfix operator: Precedence groups related to each other using lower group names and higher group
names must fit into a single relational hierarchy, but they don’t have to form a linear
postfix operator operator name hierarchy. This means it is possible to have precedence groups with undefined relative
A postfix operator is a unary operator that is written immediately after its operand, precedence. Operators from those precedence groups can’t be used next to each other
such as the postfix forced-unwrap operator (!) in the expression a!. without grouping parentheses.

Swift defines numerous precedence groups to go along with the operators provided by
As with prefix operators, postfix operator declarations don’t specify a precedence the standard library. For example, the addition (+) and subtraction (-) operators belong
level. Postfix operators are nonassociative. to the AdditionPrecedence group, and the multiplication (*) and division (/) operators
belong to the MultiplicationPrecedence group. For a complete list of operators and
After declaring a new operator, you implement it by declaring a static method that has precedence groups provided by the Swift standard library, see Swift Standard Library
the same name as the operator. The static method is a member of one of the types Operators Reference.
whose values the operator takes as an argument—for example, an operator that
multiplies a Double by an Int is implemented as a static method on either the Double The associativity of an operator specifies how a sequence of operators with the same
or Int structure. If you’re implementing a prefix or postfix operator, you must also precedence level are grouped together in the absence of grouping parentheses. You
mark that method declaration with the corresponding prefix or postfix declaration specify the associativity of an operator by writing one of the context-sensitive
modifier. To see an example of how to create and implement a new operator, see keywords left, right, or none—if your omit the associativity, the default is none.
Custom Operators. Operators that are left-associative group left-to-right. For example, the subtraction
operator (-) is left-associative, so the expression 4 - 5 - 6 is grouped as (4 - 5) -
G R A M M A R O F A N O P E R AT O R D E C L A R AT I O N 6 and evaluates to -7. Operators that are right-associative group right-to-left, and
operator-declaration → prefix-operator-declaration postfix-operator-declaration operators that are specified with an associativity of none don’t associate at all.
infix-operator-declaration Nonassociative operators of the same precedence level can’t appear adjacent to each
prefix-operator-declaration → prefixoperatoroperator to other. For example, the < operator has an associativity of none, which means 1 < 2
postfix-operator-declaration → postfixoperatoroperator < 3 is not a valid expression.
infix-operator-declaration → infixoperatoroperatorinfix-operator-groupopt
infix-operator-group → :precedence-group-name

233
The assignment of a precedence group specifies the precedence of an operator when class member can’t be overridden in any subclass. For an example of how to
used in an operation that includes optional chaining. When set to true, an operator in use the final attribute, see Preventing Overrides.
the corresponding precedence group uses the same grouping rules during optional
chaining as the assignment operators from the standard library. Otherwise, when set lazy
to false or omitted, operators in the precedence group follows the same optional Apply this modifier to a stored variable property of a class or structure to
chaining rules as operators that don’t perform assignment. indicate that the property’s initial value is calculated and stored at most once,
when the property is first accessed. For an example of how to use the lazy
G R A M M A R O F A P R E C E D E N C E G R O U P D E C L A R AT I O N modifier, see Lazy Stored Properties.
precedence-group-declaration → precedencegroupprecedence-group- optional
name{precedence-group-attributesopt}
Apply this modifier to a protocol’s property, method, or subscript members to
precedence-group-attributes → precedence-group-attributeprecedence-group-
indicate that a conforming type isn’t required to implement those members.
attributesopt
precedence-group-attribute → precedence-group-relation
You can apply the optional modifier only to protocols that are marked with
precedence-group-attribute → precedence-group-assignment
the objc attribute. As a result, only class types can adopt and conform to a
precedence-group-attribute → precedence-group-associativity
protocol that contains optional member requirements. For more information
precedence-group-relation → higherThan:precedence-group-names
about how to use the optional modifier and for guidance about how to access
precedence-group-relation → lowerThan:precedence-group-names
optional protocol members—for example, when you’re not sure whether a
precedence-group-assignment → assignment:boolean-literal
conforming type implements them—see Optional Protocol Requirements.
precedence-group-associativity → associativity:left
precedence-group-associativity → associativity:right required
precedence-group-associativity → associativity:none
Apply this modifier to a designated or convenience initializer of a class to
precedence-group-names → precedence-group-name precedence-group-
indicate that every subclass must implement that initializer. The subclass’s
name,precedence-group-names
implementation of that initializer must also be marked with the required
precedence-group-name → identifier
modifier.

Declaration Modifiers unowned

Declaration modifiers are keywords or context-sensitive keywords that modify the Apply this modifier to a stored variable, constant, or stored property to indicate
behavior or meaning of a declaration. You specify a declaration modifier by writing the that the variable or property has an unowned reference to the object stored as
appropriate keyword or context-sensitive keyword between a declaration’s attributes its value. If you try to access the variable or property after the object has been
(if any) and the keyword that introduces the declaration. deallocated, a runtime error is raised. Like a weak reference, the type of the
property or value must be a class type; unlike a weak reference, the type is
dynamic nonoptional. For an example and more information about the unowned
Apply this modifier to any member of a class that can be represented by modifier, see Unowned References.
Objective-C. When you mark a member declaration with the dynamic modifier,
unowned(safe)
access to that member is always dynamically dispatched using the Objective-
C runtime. Access to that member is never inlined or devirtualized by the An explicit spelling of unowned.
compiler.
unowned(unsafe)

Because declarations marked with the dynamic modifier are dispatched using Apply this modifier to a stored variable, constant, or stored property to indicate
the Objective-C runtime, they’re implicitly marked with the objc attribute. that the variable or property has an unowned reference to the object stored as
its value. If you try to access the variable or property after the object has been
final deallocated, you’ll access the memory at the location where the object used to
Apply this modifier to a class or to a property, method, or subscript member of be, which is a memory-unsafe operation. Like a weak reference, the type of
a class. It’s applied to a class to indicate that the class can’t be subclassed. the property or value must be a class type; unlike a weak reference, the type
It’s applied to a property, method, or subscript of a class to indicate that a

234
 is nonoptional. For an example and more information about the unowned Each access-level modifier above optionally accepts a single argument, which
modifier, see Unowned References. consists of the set keyword enclosed in parentheses (for instance, private(set)).
Use this form of an access-level modifier when you want to specify an access level for
weak the setter of a variable or subscript that’s less than or equal to the access level of the
Apply this modifier to a stored variable or stored variable property to indicate variable or subscript itself, as discussed in Getters and Setters.
that the variable or property has a weak reference to the object stored as its
value. The type of the variable or property must be an optional class type. If G R A M M A R O F A D E C L A R AT I O N M O D I F I E R
you access the variable or property is accessed after the object has been
declaration-modifier → class convenience dynamic final infix lazy optional
deallocated, its value is nil. For an example and more information about the
override postfix prefix required static unowned unowned(safe)
weak modifier, see Weak References.
unowned(unsafe) weak
declaration-modifier → access-level-modifier
Access Control Levels
declaration-modifier → mutation-modifier
Swift provides five levels of access control: open, public, internal, file private, and declaration-modifiers → declaration-modifierdeclaration-modifiersopt
private. You can mark a declaration with one of the access-level modifiers below to access-level-modifier → private private(set)
specify the declaration’s access level. Access control is discussed in detail in Access access-level-modifier → fileprivate fileprivate(set)
Control. access-level-modifier → internal internal(set)
access-level-modifier → public public(set)
open
access-level-modifier → open open(set)
Apply this modifier to a declaration to indicate the declaration can be mutation-modifier → mutating nonmutating
accessed and subclassed by code in the same module as the declaration.
Declarations marked with the open access-level modifier can also be
accessed and subclassed by code in a module that imports the module that
contains that declaration.

public
Apply this modifier to a declaration to indicate the declaration can be
accessed and subclassed by code in the same module as the declaration.
Declarations marked with the public access-level modifier can also be
accessed (but not subclassed) by code in a module that imports the module
that contains that declaration.

internal
Apply this modifier to a declaration to indicate the declaration can be
accessed only by code in the same module as the declaration. By default,
most declarations are implicitly marked with the internal access-level
modifier.

fileprivate
Apply this modifier to a declaration to indicate the declaration can be
accessed only by code in the same source file as the declaration.

private
Apply this modifier to a declaration to indicate the declaration can be
accessed only by code within the declaration’s immediate enclosing scope.

235
 Section 7 The remaining arguments can appear in any order and specify additional
information about the declaration’s lifecycle, including important milestones.

Attributes The unavailable argument indicates that the declaration isn’t

available on the specified platform.
The introduced argument indicates the first version of the
specified platform in which the declaration was introduced. It has
Attributes the following form:
introduced: version number
Attributes provide more information about a declaration or type. There are two kinds of The version number consists of one or more positive integers,
attributes in Swift, those that apply to declarations and those that apply to types. separated by periods.
The deprecated argument indicates the first version of the
You specify an attribute by writing the @ symbol followed by the attribute’s name and
specified platform in which the declaration was deprecated. It has
any arguments that the attribute accepts:
the following form:
@attribute name deprecated: version number
@attribute name(attribute arguments) The optional version number consists of one or more positive
Some declaration attributes accept arguments that specify more information about the integers, separated by periods. Omitting the version number
attribute and how it applies to a particular declaration. These attribute arguments are indicates that the declaration is currently deprecated, without
enclosed in parentheses, and their format is defined by the attribute they belong to. giving any information about when the deprecation occurred. If
you omit the version number, omit the colon (:) as well.
Declaration Attributes The obsoleted argument indicates the first version of the specified
You can apply a declaration attribute to declarations only. platform in which the declaration was obsoleted. When a
declaration is obsoleted, it’s removed from the specified platform
available and can no longer be used. It has the following form:
Apply this attribute to any declaration to indicate the declaration’s lifecycle obsoleted: version number
relative to certain platforms and operating system versions. The version number consists of one or more positive integers,
separated by periods.
The available attribute always appears with a list of two or more comma- The message argument is used to provide a textual message that’s
separated attribute arguments. These arguments begin with one of the displayed by the compiler when emitting a warning or error about
following platform names: the use of a deprecated or obsoleted declaration. It has the
iOS following form:
iOSApplicationExtension message: message
The message consists of a string literal.
macOS
macOSApplicationExtension The renamed argument is used to provide a textual message that
indicates the new name for a declaration that’s been renamed. The
watchOS
new name is displayed by the compiler when emitting an error
watchOSApplicationExtension about the use of a renamed declaration. It has the following form:
tvOS renamed: new name
tvOSApplicationExtension The new name consists of a string literal.
You can also use an asterisk (*) to indicate the availability of the declaration You can use the renamed argument in conjunction with the
on all of the platform names listed above. unavailable argument and a type alias declaration to indicate
to clients of your code that a declaration has been renamed. For

236
 example, this is useful when the name of a declaration is Apply this attribute to any declaration that can be represented in Objective-C
changed between releases of a framework or library. —for example, non-nested classes, protocols, nongeneric enumerations
// First release (constrained to integer raw-value types), properties and methods (including
getters and setters) of classes and protocols, initializers, deinitializers, and
protocol MyProtocol {
subscripts. The objc attribute tells the compiler that a declaration is available
// protocol definition to use in Objective-C code.
}
// Subsequent release renames MyProtocol Classes marked with the objc attribute must inherit from a class defined in
Objective-C. If you apply the objc attribute to a class or protocol, it’s implicitly
protocol MyRenamedProtocol {
applied to the Objective-C compatible members of that class or protocol. The
// protocol definition compiler also implicitly adds the objc attribute to a class that inherits from
} another class marked with the objc attribute or a class defined in Objective-C.
Protocols marked with the objc attribute can’t inherit from protocols that
@available(*, unavailable, renamed: aren’t.
"MyRenamedProtocol")
If you apply the objc attribute to an enumeration, each enumeration case is
typealias MyProtocol = MyRenamedProtocol
exposed to Objective-C code as the concatenation of the enumeration name
You can apply multiple available attributes on a single declaration to specify and the case name. The first letter of the case name is capitalized. For
the declaration’s availability on different platforms. The compiler uses an example, a case named venus in a Swift Planet enumeration is exposed to
available attribute only when the attribute specifies a platform that matches Objective-C code as a case named PlanetVenus.
the current target platform.
The objc attribute optionally accepts a single attribute argument, which
If an available attribute only specifies an introduced argument in addition to consists of an identifier. Use this attribute when you want to expose a different
a platform name argument, the following shorthand syntax can be used name to Objective-C for the entity the objc attribute applies to. You can use
instead: this argument to name classes, enumerations, enumeration cases, protocols,
methods, getters, setters, and initializers. The example below exposes the
@available(platform name version number, *)
getter for the enabled property of the ExampleClass to Objective-C code as
The shorthand syntax for available attributes allows for availability for isEnabled rather than just as the name of the property itself.
multiple platforms to be expressed concisely. Although the two forms are
functionally equivalent, the shorthand form is preferred whenever possible. @objc
class ExampleClass: NSObject {
@available(iOS 10.0, macOS 10.12, *)
var enabled: Bool {
class MyClass {
@objc(isEnabled) get {
// class definition
// Return the appropriate value
}
}
discardableResult
}
Apply this attribute to a function or method declaration to suppress the
}
compiler warning when the function or method that returns a value is called
without using its result. nonobjc
Apply this attribute to a method, property, subscript, or initializer declaration to
GKInspectable
suppress an implicit objc attribute. The nonobjc attribute tells the compiler to
Apply this attribute to expose a custom GameplayKit component property to make the declaration unavailable in Objective-C code, even though it is
the SpriteKit editor UI. possible to represent it in Objective-C.
objc

237
 You use the nonobjc attribute to resolve circularity for bridging methods in a UIApplicationMain
class marked with the objc attribute, and to allow overloading of methods and Apply this attribute to a class to indicate that it is the application delegate.
initializers in a class marked with the objc attribute. Using this attribute is equivalent to calling the UIApplicationMain function
and passing this class’s name as the name of the delegate class.
A method marked with the nonobjc attribute cannot override a method marked
with the objc attribute. However, a method marked with the objc attribute can If you do not use this attribute, supply a main.swift file with code at the top
override a method marked with the nonobjc attribute. Similarly, a method level that calls the UIApplicationMain(_:_:_:) function. For example, if your
marked with the nonobjc attribute cannot satisfy a protocol requirement for a app uses a custom subclass of UIApplication as its principal class, call the
method marked with the objc attribute. UIApplicationMain(_:_:_:) function instead of using this attribute.

NSApplicationMain Declaration Attributes Used by Interface Builder

Apply this attribute to a class to indicate that it is the application delegate. Interface Builder attributes are declaration attributes used by Interface Builder to
Using this attribute is equivalent to calling the NSApplicationMain(_:_:) synchronize with Xcode. Swift provides the following Interface Builder attributes:
function. IBAction, IBOutlet, IBDesignable, and IBInspectable. These attributes are
conceptually the same as their Objective-C counterparts.
If you do not use this attribute, supply a main.swift file with code at the top
level that calls the NSApplicationMain(_:_:) function as follows:
You apply the IBOutlet and IBInspectable attributes to property declarations of a
import AppKit class. You apply the IBAction attribute to method declarations of a class and the
IBDesignable attribute to class declarations.
NSApplicationMain(CommandLine.argc, CommandLine.unsafeArgv)
NSCopying
Both the IBAction and IBOutlet attributes imply the objc attribute.
Apply this attribute to a stored variable property of a class. This attribute
causes the property’s setter to be synthesized with a copy of the property’s Type Attributes
value—returned by the copyWithZone(_:) method—instead of the value of
You can apply type attributes to types only.
the property itself. The type of the property must conform to the NSCopying
protocol.
autoclosure

The NSCopying attribute behaves in a way similar to the Objective-C copy This attribute is used to delay the evaluation of an expression by automatically
property attribute. wrapping that expression in a closure with no arguments. Apply this attribute
to a parameter’s type in a method or function declaration, for a parameter of a
NSManaged function type that takes no arguments and that returns a value of the type of
Apply this attribute to an instance method or stored variable property of a the expression. For an example of how to use the autoclosure attribute, see
class that inherits from NSManagedObject to indicate that Core Data Autoclosures and Function Type.
dynamically provides its implementation at runtime, based on the associated
convention
entity description. For a property marked with the NSManaged attribute, Core
Data also provides the storage at runtime. Applying this attribute also implies Apply this attribute to the type of a function to indicate its calling conventions.
the objc attribute.
The convention attribute always appears with one of the attribute arguments
testable below.
Apply this attribute to import declarations for modules compiled with testing
enabled to access any entities marked with the internal access-level The swift argument is used to indicate a Swift function reference.
modifier as if they were declared with the public access-level modifier. Tests This is the standard calling convention for function values in Swift.
can also access classes and class members that are marked with the The block argument is used to indicate an Objective-C compatible
internal or public access-level modifier as if they were declared with the block reference. The function value is represented as a reference
open access-level modifier. to the block object, which is an id-compatible Objective-C object

238
 that embeds its invocation function within the object. The
invocation function uses the C calling convention.
The c argument is used to indicate a C function reference. The
function value carries no context and uses the C calling
convention.
A function with C function calling conventions can be used as a function with
Objective-C block calling conventions, and a function with Objective-C block
calling conventions can be used as a function with Swift function calling
conventions. However, only nongeneric global functions, and local functions or
closures that don’t capture any local variables, can be used as a function with
C function calling conventions.

escaping
Apply this attribute to a parameter’s type in a method or function declaration
to indicate that the parameter’s value can be stored for later execution. This
means that the value is allowed to outlive the lifetime of the call. Function type
parameters with the escaping type attribute require explicit use of self. for
properties or methods. For an example of how to use the escaping attribute,
see Escaping Closures.

G R A M M A R O F A N AT T R I B U T E

attribute → @attribute-nameattribute-argument-clauseopt
attribute-name → identifier
attribute-argument-clause → (balanced-tokensopt)
attributes → attributeattributesopt
balanced-tokens → balanced-tokenbalanced-tokensopt
balanced-token → (balanced-tokensopt)
balanced-token → [balanced-tokensopt]
balanced-token → {balanced-tokensopt}
balanced-token → Any identifier, keyword, literal, or operator
balanced-token → Any punctuation except (, ), [, ], {, or }

239
 Section 8 For example, the following code iterates through the closed range 1...3, ignoring the
current value of the range on each iteration of the loop:

Patterns for _ in 1...3 {

// Do something three times.
}

Patterns G R A M M A R O F A W I L D C A R D PAT T E R N

wildcard-pattern → _
A pattern represents the structure of a single value or a composite value. For
example, the structure of a tuple (1, 2) is a comma-separated list of two elements. Identifier Pattern
Because patterns represent the structure of a value rather than any one particular
An identifier pattern matches any value and binds the matched value to a variable or
value, you can match them with a variety of values. For instance, the pattern (x, y)
constant name. For example, in the following constant declaration, someValue is an
matches the tuple (1, 2) and any other two-element tuple. In addition to matching a
identifier pattern that matches the value 42 of type Int:
pattern with a value, you can extract part or all of a composite value and bind each
part to a constant or variable name. let someValue = 42

In Swift, there are two basic kinds of patterns: those that successfully match any kind When the match succeeds, the value 42 is bound (assigned) to the constant name
someValue.
of value, and those that may fail to match a specified value at runtime.

The first kind of pattern is used for destructuring values in simple variable, constant, When the pattern on the left-hand side of a variable or constant declaration is an
and optional bindings. These include wildcard patterns, identifier patterns, and any identifier pattern, the identifier pattern is implicitly a subpattern of a value-binding
value binding or tuple patterns containing them. You can specify a type annotation for pattern.
these patterns to constrain them to match only values of a certain type.
G R A M M A R O F A N I D E N T I F I E R PAT T E R N

The second kind of pattern is used for full pattern matching, where the values you’re identifier-pattern → identifier
trying to match against may not be there at runtime. These include enumeration case
patterns, optional patterns, expression patterns, and type-casting patterns. You use Value-Binding Pattern
these patterns in a case label of a switch statement, a catch clause of a do
A value-binding pattern binds matched values to variable or constant names. Value-
statement, or in the case condition of an if, while, guard, or for-in statement.
binding patterns that bind a matched value to the name of a constant begin with the
let keyword; those that bind to the name of variable begin with the var keyword.
G R A M M A R O F A PAT T E R N

pattern → wildcard-patterntype-annotationopt Identifiers patterns within a value-binding pattern bind new named variables or
pattern → identifier-patterntype-annotationopt constants to their matching values. For example, you can decompose the elements of
pattern → value-binding-pattern a tuple and bind the value of each element to a corresponding identifier pattern.
pattern → tuple-patterntype-annotationopt
pattern → enum-case-pattern let point = (3, 2)
pattern → optional-pattern switch point {
pattern → type-casting-pattern // Bind x and y to the elements of point.
pattern → expression-pattern
case let (x, y):
print("The point is at (\(x), \(y)).")
Wildcard Pattern
}
A wildcard pattern matches and ignores any value and consists of an underscore (_).
Use a wildcard pattern when you don’t care about the values being matched against. // Prints "The point is at (3, 2)."

240
In the example above, let distributes to each identifier pattern in the tuple pattern (x, An enumeration case pattern matches a case of an existing enumeration type.
y). Because of this behavior, the switch cases case let (x, y): and case (let x, Enumeration case patterns appear in switch statement case labels and in the case
let y): match the same values. conditions of if, while, guard, and for-in statements.

G R A M M A R O F A VA L U E - B I N D I N G PAT T E R N If the enumeration case you’re trying to match has any associated values, the
value-binding-pattern → varpattern letpattern corresponding enumeration case pattern must specify a tuple pattern that contains
one element for each associated value. For an example that uses a switch statement
to match enumeration cases containing associated values, see Associated Values.
Tuple Pattern
A tuple pattern is a comma-separated list of zero or more patterns, enclosed in G R A M M A R O F A N E N U M E R AT I O N C A S E PAT T E R N
parentheses. Tuple patterns match values of corresponding tuple types.
enum-case-pattern → type-identifieropt.enum-case-nametuple-patternopt
You can constrain a tuple pattern to match certain kinds of tuple types by using type
annotations. For example, the tuple pattern (x, y): (Int, Int) in the constant Optional Pattern
declaration let (x, y): (Int, Int) = (1, 2) matches only tuple types in which An optional pattern matches values wrapped in a some(Wrapped) case of an
both elements are of type Int. Optional<Wrapped> enumeration. Optional patterns consist of an identifier pattern
followed immediately by a question mark and appear in the same places as
When a tuple pattern is used as the pattern in a for-in statement or in a variable or enumeration case patterns.
constant declaration, it can contain only wildcard patterns, identifier patterns, optional
patterns, or other tuple patterns that contain those. For example, the following code Because optional patterns are syntactic sugar for Optional enumeration case
isn’t valid because the element 0 in the tuple pattern (x, 0) is an expression pattern: patterns, the following are equivalent:

let points = [(0, 0), (1, 0), (1, 1), (2, 0), (2, 1)] let someOptional: Int? = 42
// This code isn't valid. // Match using an enumeration case pattern.
for (x, 0) in points { if case .some(let x) = someOptional {
/* ... */ print(x)
} }

The parentheses around a tuple pattern that contains a single element have no effect.
The pattern matches values of that single element’s type. For example, the following // Match using an optional pattern.
are equivalent: if case let x? = someOptional {
print(x)
let a = 2 // a: Int = 2
}
let (a) = 2 // a: Int = 2
let (a): Int = 2 // a: Int = 2 The optional pattern provides a convenient way to iterate over an array of optional
values in a for-in statement, executing the body of the loop only for non-nil
G R A M M A R O F A T U P L E PAT T E R N elements.
tuple-pattern → (tuple-pattern-element-listopt)
tuple-pattern-element-list → tuple-pattern-element tuple-pattern-element,tuple- let arrayOfOptionalInts: [Int?] = [nil, 2, 3, nil, 5]
pattern-element-list // Match only non-nil values.
tuple-pattern-element → pattern identifier:pattern for case let number? in arrayOfOptionalInts {
print("Found a \(number)")
Enumeration Case Pattern }
// Found a 2

241
 // Found a 3 case (0, 0):
// Found a 5 print("(0, 0) is at the origin.")

G R A M M A R O F A N O P T I O N A L PAT T E R N case (-2...2, -2...2):

print("(\(point.0), \(point.1)) is near the origin.")
optional-pattern → identifier-pattern?
default:

Type-Casting Patterns print("The point is at (\(point.0), \(point.1)).")

}
There are two type-casting patterns, the is pattern and the as pattern. The is pattern
appears only in switch statement case labels. The is and as patterns have the // Prints "(1, 2) is near the origin."
following form: You can overload the ~= operator to provide custom expression matching behavior.
For example, you can rewrite the above example to compare the point expression
is type
with a string representations of points.
pattern as type
The is pattern matches a value if the type of that value at runtime is the same as the // Overload the ~= operator to match a string with an integer.
type specified in the right-hand side of the is pattern—or a subclass of that type. The
func ~=(pattern: String, value: Int) -> Bool {
is pattern behaves like the is operator in that they both perform a type cast but
discard the returned type. return pattern == "\(value)"
}
The as pattern matches a value if the type of that value at runtime is the same as the switch point {
type specified in the right-hand side of the as pattern—or a subclass of that type. If the case ("0", "0"):
match succeeds, the type of the matched value is cast to the pattern specified in the
print("(0, 0) is at the origin.")
right-hand side of the as pattern.
default:

For an example that uses a switch statement to match values with is and as patterns, print("The point is at (\(point.0), \(point.1)).")
see Type Casting for Any and AnyObject. }
// Prints "The point is at (1, 2)."
G R A M M A R O F A T Y P E C A S T I N G PAT T E R N
G R A M M A R O F A N E X P R E S S I O N PAT T E R N
type-casting-pattern → is-pattern as-pattern
is-pattern → istype expression-pattern → expression
as-pattern → patternastype

Expression Pattern
An expression pattern represents the value of an expression. Expression patterns
appear only in switch statement case labels.

The expression represented by the expression pattern is compared with the value of
an input expression using the Swift standard library ~= operator. The matches
succeeds if the ~= operator returns true. By default, the ~= operator compares two
values of the same type using the == operator. It can also match a value with a range
of values, by checking whether the value is contained within the range, as the
following example shows.

let point = (1, 2)

switch point {

242
 return x
Section 9 }

Generic Parameters and Because Int and Double, for example, both conform to the Comparable protocol, this
function accepts arguments of either type. In contrast with generic types, you don’t
specify a generic argument clause when you use a generic function or initializer. The
type arguments are instead inferred from the type of the arguments passed to the
function or initializer.
Generic Parameters and Arguments
simpleMax(17, 42) // T is inferred to be Int
This chapter describes parameters and arguments for generic types, functions, and simpleMax(3.14159, 2.71828) // T is inferred to be Double
initializers. When you declare a generic type, function, or initializer, you specify the
type parameters that the generic type, function, or initializer can work with. These type Generic Where Clauses
parameters act as placeholders that are replaced by actual concrete type arguments You can specify additional requirements on type parameters and their associated
when an instance of a generic type is created or a generic function or initializer is types by including a generic where clause right before the opening curly brace of a
called. type or function’s body. A generic where clause consists of the where keyword,
followed by a comma-separated list of one or more requirements.
For an overview of generics in Swift, see Generics.
where requirements
Generic Parameter Clause The requirements in a generic where clause specify that a type parameter inherits
A generic parameter clause specifies the type parameters of a generic type or from a class or conforms to a protocol or protocol composition. Although the generic
function, along with any associated constraints and requirements on those where clause provides syntactic sugar for expressing simple constraints on type
parameters. A generic parameter clause is enclosed in angle brackets (<>) and has parameters (for instance, <T: Comparable> is equivalent to <T> where T: Comparable
the following form: and so on), you can use it to provide more complex constraints on type parameters
and their associated types. For instance, you can constrain the associated types of
<generic parameter list> type parameters to conform to protocols. For example, <S: Sequence> where
The generic parameter list is a comma-separated list of generic parameters, each of S.Iterator.Element: Equatable specifies that S conforms to the Sequence protocol
which has the following form: and that the associated type S.Iterator.Element conforms to the Equatable
protocol. This constraint ensures that each element of the sequence is equatable.
type parameter: constraint
A generic parameter consists of a type parameter followed by an optional constraint. A You can also specify the requirement that two types be identical, using the ==
type parameter is simply the name of a placeholder type (for instance, T, U, V, Key, operator. For example, <S1: Sequence, S2: Sequence> where S1.Iterator.Element
Value, and so on). You have access to the type parameters (and any of their == S2.Iterator.Element expresses the constraints that S1 and S2 conform to the
associated types) in the rest of the type, function, or initializer declaration, including in Sequence protocol and that the elements of both sequences must be of the same type.
the signature of the function or initializer.
Any type argument substituted for a type parameter must meet all the constraints and
The constraint specifies that a type parameter inherits from a specific class or requirements placed on the type parameter.
conforms to a protocol or protocol composition. For instance, in the generic function
below, the generic parameter T: Comparable indicates that any type argument You can overload a generic function or initializer by providing different constraints,
substituted for the type parameter T must conform to the Comparable protocol. requirements, or both on the type parameters. When you call an overloaded generic
function or initializer, the compiler uses these constraints to resolve which overloaded
func simpleMax<T: Comparable>(_ x: T, _ y: T) -> T { function or initializer to invoke.
if x < y {
For more information about generic where clauses and to see an example of one in a
return y
generic function declaration, see Generic Where Clauses.
}

243
 G R A M M A R O F A G E N E R I C PA R A M E T E R C L A U S E let arrayOfArrays: Array<Array<Int>> = [[1, 2, 3], [4, 5, 6], [7,
8, 9]]
generic-parameter-clause → <generic-parameter-list>
generic-parameter-list → generic-parameter generic-parameter,generic- As mentioned in Generic Parameter Clause, you don’t use a generic argument clause
parameter-list to specify the type arguments of a generic function or initializer.
generic-parameter → type-name
GRAMMAR OF A GENERIC ARGUMENT CLAUSE
generic-parameter → type-name:type-identifier
generic-parameter → type-name:protocol-composition-type generic-argument-clause → <generic-argument-list>
generic-where-clause → whererequirement-list generic-argument-list → generic-argument generic-argument,generic-argument-list
requirement-list → requirement requirement,requirement-list generic-argument → type
requirement → conformance-requirement same-type-requirement
conformance-requirement → type-identifier:type-identifier
conformance-requirement → type-identifier:protocol-composition-type
same-type-requirement → type-identifier==type

Generic Argument Clause

A generic argument clause specifies the type arguments of a generic type. A generic
argument clause is enclosed in angle brackets (<>) and has the following form:

<generic argument list>

The generic argument list is a comma-separated list of type arguments. A type
argument is the name of an actual concrete type that replaces a corresponding type
parameter in the generic parameter clause of a generic type. The result is a
specialized version of that generic type. The example below shows a simplified
version of the Swift standard library’s generic dictionary type.

struct Dictionary<Key: Hashable, Value>: Collection,

ExpressibleByDictionaryLiteral {
/* ... */
}

The specialized version of the generic Dictionary type, Dictionary<String, Int> is

formed by replacing the generic parameters Key: Hashable and Value with the
concrete type arguments String and Int. Each type argument must satisfy all the
constraints of the generic parameter it replaces, including any additional requirements
specified in a generic where clause. In the example above, the Key type parameter is
constrained to conform to the Hashable protocol and therefore String must also
conform to the Hashable protocol.

You can also replace a type parameter with a type argument that is itself a specialized
version of a generic type (provided it satisfies the appropriate constraints and
requirements). For example, you can replace the type parameter Element in
Array<Element> with a specialized version of an array, Array<Int>, to form an array
whose elements are themselves arrays of integers.

244
 Section 10 identifier-character → identifier-head
identifier-characters → identifier-characteridentifier-charactersopt

Summary of the Grammar implicit-parameter-name → $decimal-digits

GRAMMAR OF A LITERAL

literal → numeric-literal string-literal boolean-literal nil-literal

numeric-literal → -optinteger-literal -optfloating-point-literal
Summary of the Grammar boolean-literal → true false
nil-literal → nil
Lexical Structure GRAMMAR OF AN INTEGER LITERAL

GRAMMAR OF AN IDENTIFIER integer-literal → binary-literal

integer-literal → octal-literal
identifier → identifier-headidentifier-charactersopt integer-literal → decimal-literal
identifier → `identifier-headidentifier-charactersopt` integer-literal → hexadecimal-literal
identifier → implicit-parameter-name binary-literal → 0bbinary-digitbinary-literal-charactersopt
identifier-list → identifier identifier,identifier-list binary-digit → Digit 0 or 1
identifier-head → Upper- or lowercase letter A through Z binary-literal-character → binary-digit _
identifier-head → _ binary-literal-characters → binary-literal-characterbinary-literal-charactersopt
identifier-head → U+00A8, U+00AA, U+00AD, U+00AF, U+00B2–U+00B5, or octal-literal → 0ooctal-digitoctal-literal-charactersopt
U+00B7–U+00BA octal-digit → Digit 0 through 7
identifier-head → U+00BC–U+00BE, U+00C0–U+00D6, U+00D8–U+00F6, or octal-literal-character → octal-digit _
U+00F8–U+00FF octal-literal-characters → octal-literal-characteroctal-literal-charactersopt
identifier-head → U+0100–U+02FF, U+0370–U+167F, U+1681–U+180D, or decimal-literal → decimal-digitdecimal-literal-charactersopt
U+180F–U+1DBF decimal-digit → Digit 0 through 9
identifier-head → U+1E00–U+1FFF decimal-digits → decimal-digitdecimal-digitsopt
identifier-head → U+200B–U+200D, U+202A–U+202E, U+203F–U+2040, U+2054, decimal-literal-character → decimal-digit _
or U+2060–U+206F decimal-literal-characters → decimal-literal-characterdecimal-literal-charactersopt
identifier-head → U+2070–U+20CF, U+2100–U+218F, U+2460–U+24FF, or hexadecimal-literal → 0xhexadecimal-digithexadecimal-literal-charactersopt
U+2776–U+2793 hexadecimal-digit → Digit 0 through 9, a through f, or A through F
identifier-head → U+2C00–U+2DFF or U+2E80–U+2FFF hexadecimal-literal-character → hexadecimal-digit _
identifier-head → U+3004–U+3007, U+3021–U+302F, U+3031–U+303F, or hexadecimal-literal-characters → hexadecimal-literal-characterhexadecimal-literal-
U+3040–U+D7FF charactersopt
identifier-head → U+F900–U+FD3D, U+FD40–U+FDCF, U+FDF0–U+FE1F, or
U+FE30–U+FE44 G R A M M A R O F A F L O AT I N G - P O I N T L I T E R A L
identifier-head → U+FE47–U+FFFD floating-point-literal → decimal-literaldecimal-fractionoptdecimal-exponentopt
identifier-head → U+10000–U+1FFFD, U+20000–U+2FFFD, U+30000–U+3FFFD, floating-point-literal → hexadecimal-literalhexadecimal-fractionopthexadecimal-
or U+40000–U+4FFFD exponent
identifier-head → U+50000–U+5FFFD, U+60000–U+6FFFD, U+70000–U+7FFFD, decimal-fraction → .decimal-literal
or U+80000–U+8FFFD decimal-exponent → floating-point-esignoptdecimal-literal
identifier-head → U+90000–U+9FFFD, U+A0000–U+AFFFD, U+B0000– hexadecimal-fraction → .hexadecimal-digithexadecimal-literal-charactersopt
U+BFFFD, or U+C0000–U+CFFFD hexadecimal-exponent → floating-point-psignoptdecimal-literal
identifier-head → U+D0000–U+DFFFD or U+E0000–U+EFFFD floating-point-e → e E
identifier-character → Digit 0 through 9 floating-point-p → p P
identifier-character → U+0300–U+036F, U+1DC0–U+1DFF, U+20D0–U+20FF, or sign → + -
U+FE20–U+FE2F

245
GRAMMAR OF A STRING LITERAL postfix-operator → operator
string-literal → static-string-literal interpolated-string-literal
static-string-literal → "quoted-textopt" Types
quoted-text → quoted-text-itemquoted-textopt GRAMMAR OF A TYPE
quoted-text-item → escaped-character
quoted-text-item → Any Unicode scalar value except ", \, U+000A, or U+000D type → array-type dictionary-type function-type type-identifier tuple-type optional-
interpolated-string-literal → "interpolated-textopt" type implicitly-unwrapped-optional-type protocol-composition-type metatype-
interpolated-text → interpolated-text-iteminterpolated-textopt type Any Self
interpolated-text-item → \(expression) quoted-text-item G R A M M A R O F A T Y P E A N N O TAT I O N
escaped-character → \0 \\ \t \n \r \" \'
escaped-character → \u{unicode-scalar-digits} type-annotation → :attributesoptinoutopttype
unicode-scalar-digits → Between one and eight hexadecimal digits GRAMMAR OF A TYPE IDENTIFIER
G R A M M A R O F O P E R AT O R S
type-identifier → type-namegeneric-argument-clauseopt type-namegeneric-
operator → operator-headoperator-charactersopt argument-clauseopt.type-identifier
operator → dot-operator-headdot-operator-characters type-name → identifier
operator-head → / = - + ! * % < > & | ^ ~ ? GRAMMAR OF A TUPLE TYPE
operator-head → U+00A1–U+00A7
operator-head → U+00A9 or U+00AB tuple-type → (tuple-type-element-listopt)
operator-head → U+00AC or U+00AE tuple-type-element-list → tuple-type-element tuple-type-element,tuple-type-
operator-head → U+00B0–U+00B1, U+00B6, U+00BB, U+00BF, U+00D7, or element-list
U+00F7 tuple-type-element → element-nametype-annotation type
operator-head → U+2016–U+2017 or U+2020–U+2027 element-name → identifier
operator-head → U+2030–U+203E GRAMMAR OF A FUNCTION TYPE
operator-head → U+2041–U+2053
operator-head → U+2055–U+205E function-type → attributesoptfunction-type-argument-clausethrowsopt->type
operator-head → U+2190–U+23FF function-type → attributesoptfunction-type-argument-clauserethrows->type
operator-head → U+2500–U+2775 function-type-argument-clause → ()
operator-head → U+2794–U+2BFF function-type-argument-clause → (function-type-argument-list...opt)
operator-head → U+2E00–U+2E7F function-type-argument-list → function-type-argument function-type-
operator-head → U+3001–U+3003 argument,function-type-argument-list
operator-head → U+3008–U+3030 function-type-argument → attributesoptinoutopttype argument-labeltype-annotation
operator-character → operator-head argument-label → identifier
operator-character → U+0300–U+036F G R A M M A R O F A N A R R AY T Y P E
operator-character → U+1DC0–U+1DFF
operator-character → U+20D0–U+20FF array-type → [type]
operator-character → U+FE00–U+FE0F GRAMMAR OF A DICTIONARY TYPE
operator-character → U+FE20–U+FE2F
operator-character → U+E0100–U+E01EF dictionary-type → [type:type]
operator-characters → operator-characteroperator-charactersopt GRAMMAR OF AN OPTIONAL TYPE
dot-operator-head → .
dot-operator-character → . operator-character optional-type → type?
dot-operator-characters → dot-operator-characterdot-operator-charactersopt G R A M M A R O F A N I M P L I C I T LY U N W R A P P E D O P T I O N A L T Y P E
binary-operator → operator
prefix-operator → operator implicitly-unwrapped-optional-type → type!

246
 GRAMMAR OF A PROTOCOL COMPOSITION TYPE type-casting-operator → as?type
protocol-composition-type → protocol-identifier&protocol-composition-continuation type-casting-operator → as!type
protocol-composition-continuation → protocol-identifier protocol-composition-type GRAMMAR OF A PRIMARY EXPRESSION
protocol-identifier → type-identifier
primary-expression → identifiergeneric-argument-clauseopt
G R A M M A R O F A M E TAT Y P E T Y P E primary-expression → literal-expression
metatype-type → type.Type type.Protocol primary-expression → self-expression
primary-expression → superclass-expression
G R A M M A R O F A T Y P E I N H E R I TA N C E C L A U S E primary-expression → closure-expression
type-inheritance-clause → :class-requirement,type-inheritance-list primary-expression → parenthesized-expression
type-inheritance-clause → :class-requirement primary-expression → tuple-expression
type-inheritance-clause → :type-inheritance-list primary-expression → implicit-member-expression
type-inheritance-list → type-identifier type-identifier,type-inheritance-list primary-expression → wildcard-expression
class-requirement → class primary-expression → selector-expression
primary-expression → key-path-expression
Expressions GRAMMAR OF A LITERAL EXPRESSION

GRAMMAR OF AN EXPRESSION literal-expression → literal

literal-expression → array-literal dictionary-literal playground-literal
expression → try-operatoroptprefix-expressionbinary-expressionsopt literal-expression → #file #line #column #function
expression-list → expression expression,expression-list array-literal → [array-literal-itemsopt]
GRAMMAR OF A PREFIX EXPRESSION array-literal-items → array-literal-item,opt array-literal-item,array-literal-items
array-literal-item → expression
prefix-expression → prefix-operatoroptpostfix-expression dictionary-literal → [dictionary-literal-items] [:]
prefix-expression → in-out-expression dictionary-literal-items → dictionary-literal-item,opt dictionary-literal-item,dictionary-
in-out-expression → &identifier literal-items
GRAMMAR OF A TRY EXPRESSION dictionary-literal-item → expression:expression
playground-literal →
try-operator → try try? try! #colorLiteral(red:expression,green:expression,blue:expression,alpha:e
GRAMMAR OF A BINARY EXPRESSION xpression)
playground-literal → #fileLiteral(resourceName:expression)
binary-expression → binary-operatorprefix-expression playground-literal → #imageLiteral(resourceName:expression)
binary-expression → assignment-operatortry-operatoroptprefix-expression
binary-expression → conditional-operatortry-operatoroptprefix-expression GRAMMAR OF A SELF EXPRESSION
binary-expression → type-casting-operator self-expression → self self-method-expression self-subscript-expression self-
binary-expressions → binary-expressionbinary-expressionsopt initializer-expression
G R A M M A R O F A N A S S I G N M E N T O P E R AT O R self-method-expression → self.identifier
self-subscript-expression → self[expression-list]
assignment-operator → = self-initializer-expression → self.init
G R A M M A R O F A C O N D I T I O N A L O P E R AT O R
GRAMMAR OF A SUPERCLASS EXPRESSION
conditional-operator → ?try-operatoroptexpression: superclass-expression → superclass-method-expression superclass-subscript-
G R A M M A R O F A T Y P E - C A S T I N G O P E R AT O R expression superclass-initializer-expression
superclass-method-expression → super.identifier
type-casting-operator → istype superclass-subscript-expression → super[expression-list]
type-casting-operator → astype superclass-initializer-expression → super.init

247
GRAMMAR OF A CLOSURE EXPRESSION postfix-expression → dynamic-type-expression
closure-expression → {closure-signatureoptstatementsopt} postfix-expression → subscript-expression
closure-signature → capture-listoptclosure-parameter-clausethrowsoptfunction- postfix-expression → forced-value-expression
resultoptin postfix-expression → optional-chaining-expression
closure-signature → capture-listin GRAMMAR OF A FUNCTION CALL EXPRESSION
closure-parameter-clause → () (closure-parameter-list) identifier-list
closure-parameter-list → closure-parameter closure-parameter,closure-parameter- function-call-expression → postfix-expressionfunction-call-argument-clause
list function-call-expression → postfix-expressionfunction-call-argument-
closure-parameter → closure-parameter-nametype-annotationopt clauseopttrailing-closure
closure-parameter → closure-parameter-nametype-annotation... function-call-argument-clause → () (function-call-argument-list)
closure-parameter-name → identifier function-call-argument-list → function-call-argument function-call-
capture-list → [capture-list-items] argument,function-call-argument-list
capture-list-items → capture-list-item capture-list-item,capture-list-items function-call-argument → expression identifier:expression
capture-list-item → capture-specifieroptexpression function-call-argument → operator identifier:operator
capture-specifier → weak unowned unowned(safe) unowned(unsafe) trailing-closure → closure-expression
GRAMMAR OF AN INITIALIZER EXPRESSION
GRAMMAR OF A IMPLICIT MEMBER EXPRESSION

implicit-member-expression → .identifier initializer-expression → postfix-expression.init

initializer-expression → postfix-expression.init(argument-names)
G R A M M A R O F A PA R E N T H E S I Z E D E X P R E S S I O N
GRAMMAR OF AN EXPLICIT MEMBER EXPRESSION
parenthesized-expression → (expression)
explicit-member-expression → postfix-expression.decimal-digits
GRAMMAR OF A TUPLE EXPRESSION explicit-member-expression → postfix-expression.identifiergeneric-argument-
tuple-expression → () (tuple-element,tuple-element-list) clauseopt
tuple-element-list → tuple-element tuple-element,tuple-element-list explicit-member-expression → postfix-expression.identifier(argument-names)
tuple-element → expression identifier:expression argument-names → argument-nameargument-namesopt
argument-name → identifier:
GRAMMAR OF A WILDCARD EXPRESSION
GRAMMAR OF A SELF EXPRESSION
wildcard-expression → _
postfix-self-expression → postfix-expression.self
GRAMMAR OF A SELECTOR EXPRESSION
GRAMMAR OF A DYNAMIC TYPE EXPRESSION
selector-expression → #selector(expression)
selector-expression → #selector(getter:expression) dynamic-type-expression → type(of:expression)
selector-expression → #selector(setter:expression) GRAMMAR OF A SUBSCRIPT EXPRESSION

G R A M M A R O F A K E Y- PAT H E X P R E S S I O N subscript-expression → postfix-expression[expression-list]

key-path-expression → #keyPath(expression) G R A M M A R O F A F O R C E D - VA L U E E X P R E S S I O N

GRAMMAR OF A POSTFIX EXPRESSION forced-value-expression → postfix-expression!

postfix-expression → primary-expression GRAMMAR OF AN OPTIONAL-CHAINING EXPRESSION
postfix-expression → postfix-expressionpostfix-operator
postfix-expression → function-call-expression optional-chaining-expression → postfix-expression?
postfix-expression → initializer-expression
postfix-expression → explicit-member-expression Statements
postfix-expression → postfix-self-expression G R A M M A R O F A S TAT E M E N T

248
statement → expression;opt case-label → casecase-item-list:
statement → declaration;opt case-item-list → patternwhere-clauseopt patternwhere-clauseopt,case-item-list
statement → loop-statement;opt default-label → default:
statement → branch-statement;opt where-clause → wherewhere-expression
statement → labeled-statement;opt where-expression → expression
statement → control-transfer-statement;opt
G R A M M A R O F A L A B E L E D S TAT E M E N T
statement → defer-statement;opt
statement → do-statement:opt labeled-statement → statement-labelloop-statement
statement → compiler-control-statement labeled-statement → statement-labelif-statement
statements → statementstatementsopt labeled-statement → statement-labelswitch-statement
labeled-statement → statement-labeldo-statement
G R A M M A R O F A L O O P S TAT E M E N T
statement-label → label-name:
loop-statement → for-in-statement label-name → identifier
loop-statement → while-statement
G R A M M A R O F A C O N T R O L T R A N S F E R S TAT E M E N T
loop-statement → repeat-while-statement
control-transfer-statement → break-statement
G R A M M A R O F A F O R - I N S TAT E M E N T
control-transfer-statement → continue-statement
for-in-statement → forcaseoptpatterninexpressionwhere-clauseoptcode-block control-transfer-statement → fallthrough-statement
control-transfer-statement → return-statement
G R A M M A R O F A W H I L E S TAT E M E N T
control-transfer-statement → throw-statement
while-statement → whilecondition-listcode-block
G R A M M A R O F A B R E A K S TAT E M E N T
condition-list → condition condition,condition-list
condition → expression availability-condition case-condition optional-binding- break-statement → breaklabel-nameopt
condition
G R A M M A R O F A C O N T I N U E S TAT E M E N T
case-condition → casepatterninitializer
optional-binding-condition → letpatterninitializer varpatterninitializer continue-statement → continuelabel-nameopt
G R A M M A R O F A R E P E AT- W H I L E S TAT E M E N T G R A M M A R O F A FA L LT H R O U G H S TAT E M E N T

repeat-while-statement → repeatcode-blockwhileexpression fallthrough-statement → fallthrough

G R A M M A R O F A B R A N C H S TAT E M E N T G R A M M A R O F A R E T U R N S TAT E M E N T

branch-statement → if-statement return-statement → returnexpressionopt

branch-statement → guard-statement
G R A M M A R O F A T H R O W S TAT E M E N T
branch-statement → switch-statement
throw-statement → throwexpression
G R A M M A R O F A N I F S TAT E M E N T
G R A M M A R O F A D E F E R S TAT E M E N T
if-statement → ifcondition-listcode-blockelse-clauseopt
else-clause → elsecode-block elseif-statement defer-statement → defercode-block
G R A M M A R O F A G U A R D S TAT E M E N T G R A M M A R O F A D O S TAT E M E N T

guard-statement → guardcondition-listelsecode-block do-statement → docode-blockcatch-clausesopt

catch-clauses → catch-clausecatch-clausesopt
G R A M M A R O F A S W I T C H S TAT E M E N T
catch-clause → catchpatternoptwhere-clauseoptcode-block
switch-statement → switchexpression{switch-casesopt}
G R A M M A R O F A C O M P I L E R C O N T R O L S TAT E M E N T
switch-cases → switch-caseswitch-casesopt
switch-case → case-labelstatements default-labelstatements compiler-control-statement → conditional-compilation-block

249
compiler-control-statement → line-control-statement Declarations
G R A M M A R O F A C O N D I T I O N A L C O M P I L AT I O N B L O C K G R A M M A R O F A D E C L A R AT I O N

conditional-compilation-block → if-directive-clauseelseif-directive-clausesoptelse- declaration → import-declaration

directive-clauseoptendif-directive declaration → constant-declaration
if-directive-clause → if-directivecompilation-conditionstatementsopt declaration → variable-declaration
elseif-directive-clauses → elseif-directive-clauseelseif-directive-clausesopt declaration → typealias-declaration
elseif-directive-clause → elseif-directivecompilation-conditionstatementsopt declaration → function-declaration
else-directive-clause → else-directivestatementsopt declaration → enum-declaration
if-directive → #if declaration → struct-declaration
elseif-directive → #elseif declaration → class-declaration
else-directive → #else declaration → protocol-declaration
endif-directive → #endif declaration → initializer-declaration
compilation-condition → platform-condition declaration → deinitializer-declaration
compilation-condition → identifier declaration → extension-declaration
compilation-condition → boolean-literal declaration → subscript-declaration
compilation-condition → (compilation-condition) declaration → operator-declaration
compilation-condition → !compilation-condition declaration → precedence-group-declaration
compilation-condition → compilation-condition&&compilation-condition declarations → declarationdeclarationsopt
compilation-condition → compilation-condition||compilation-condition
platform-condition → os(operating-system) G R A M M A R O F A T O P - L E V E L D E C L A R AT I O N

platform-condition → arch(architecture) top-level-declaration → statementsopt

platform-condition → swift(>=swift-version)
operating-system → macOS iOS watchOS tvOS GRAMMAR OF A CODE BLOCK

architecture → i386 x86_64 arm arm64 code-block → {statementsopt}

swift-version → decimal-digits.decimal-digits
G R A M M A R O F A N I M P O R T D E C L A R AT I O N
G R A M M A R O F A L I N E C O N T R O L S TAT E M E N T
import-declaration → attributesoptimportimport-kindoptimport-path
line-control-statement → #sourceLocation(file:file-name,line:line-number) import-kind → typealias struct class enum protocol var func
line-control-statement → #sourceLocation() import-path → import-path-identifier import-path-identifier.import-path
line-number → A decimal integer greater than zero import-path-identifier → identifier operator
file-name → static-string-literal
G R A M M A R O F A C O N S TA N T D E C L A R AT I O N
G R A M M A R O F A N AVA I L A B I L I T Y C O N D I T I O N
constant-declaration → attributesoptdeclaration-modifiersoptletpattern-initializer-list
availability-condition → #available(availability-arguments) pattern-initializer-list → pattern-initializer pattern-initializer,pattern-initializer-list
availability-arguments → availability-argument availability-argument,availability- pattern-initializer → patterninitializeropt
arguments initializer → =expression
availability-argument → platform-nameplatform-version
availability-argument → * G R A M M A R O F A VA R I A B L E D E C L A R AT I O N

platform-name → iOS iOSApplicationExtension variable-declaration → variable-declaration-headpattern-initializer-list

platform-name → macOS macOSApplicationExtension variable-declaration → variable-declaration-headvariable-nametype-
platform-name → watchOS annotationcode-block
platform-name → tvOS variable-declaration → variable-declaration-headvariable-nametype-
platform-version → decimal-digits annotationgetter-setter-block
platform-version → decimal-digits.decimal-digits
platform-version → decimal-digits.decimal-digits.decimal-digits

250
variable-declaration → variable-declaration-headvariable-nametype- default-argument-clause → =expression
annotationgetter-setter-keyword-block
G R A M M A R O F A N E N U M E R AT I O N D E C L A R AT I O N
variable-declaration → variable-declaration-headvariable-nameinitializerwillSet-
didSet-block enum-declaration → attributesoptaccess-level-modifieroptunion-style-enum
variable-declaration → variable-declaration-headvariable-nametype- enum-declaration → attributesoptaccess-level-modifieroptraw-value-style-enum
annotationinitializeroptwillSet-didSet-block union-style-enum → indirectoptenumenum-namegeneric-parameter-clauseopttype-
variable-declaration-head → attributesoptdeclaration-modifiersoptvar inheritance-clauseoptgeneric-where-clauseopt{union-style-enum-membersopt}
variable-name → identifier union-style-enum-members → union-style-enum-memberunion-style-enum-
getter-setter-block → code-block membersopt
getter-setter-block → {getter-clausesetter-clauseopt} union-style-enum-member → declaration union-style-enum-case-clause compiler-
getter-setter-block → {setter-clausegetter-clause} control-statement
getter-clause → attributesoptmutation-modifieroptgetcode-block union-style-enum-case-clause → attributesoptindirectoptcaseunion-style-enum-
setter-clause → attributesoptmutation-modifieroptsetsetter-nameoptcode-block case-list
setter-name → (identifier) union-style-enum-case-list → union-style-enum-case union-style-enum-
getter-setter-keyword-block → {getter-keyword-clausesetter-keyword-clauseopt} case,union-style-enum-case-list
getter-setter-keyword-block → {setter-keyword-clausegetter-keyword-clause} union-style-enum-case → enum-case-nametuple-typeopt
getter-keyword-clause → attributesoptmutation-modifieroptget enum-name → identifier
setter-keyword-clause → attributesoptmutation-modifieroptset enum-case-name → identifier
willSet-didSet-block → {willSet-clausedidSet-clauseopt} raw-value-style-enum → enumenum-namegeneric-parameter-clauseopttype-
willSet-didSet-block → {didSet-clausewillSet-clauseopt} inheritance-clausegeneric-where-clauseopt{raw-value-style-enum-members}
willSet-clause → attributesoptwillSetsetter-nameoptcode-block raw-value-style-enum-members → raw-value-style-enum-memberraw-value-style-
didSet-clause → attributesoptdidSetsetter-nameoptcode-block enum-membersopt
raw-value-style-enum-member → declaration raw-value-style-enum-case-clause
G R A M M A R O F A T Y P E A L I A S D E C L A R AT I O N
compiler-control-statement
typealias-declaration → attributesoptaccess-level-modifieropttypealiastypealias- raw-value-style-enum-case-clause → attributesoptcaseraw-value-style-enum-case-
namegeneric-parameter-clauseopttypealias-assignment list
typealias-name → identifier raw-value-style-enum-case-list → raw-value-style-enum-case raw-value-style-
typealias-assignment → =type enum-case,raw-value-style-enum-case-list
raw-value-style-enum-case → enum-case-nameraw-value-assignmentopt
G R A M M A R O F A F U N C T I O N D E C L A R AT I O N
raw-value-assignment → =raw-value-literal
function-declaration → function-headfunction-namegeneric-parameter- raw-value-literal → numeric-literal static-string-literal boolean-literal
clauseoptfunction-signaturegeneric-where-clauseoptfunction-bodyopt
G R A M M A R O F A S T R U C T U R E D E C L A R AT I O N
function-head → attributesoptdeclaration-modifiersoptfunc
function-name → identifier operator struct-declaration → attributesoptaccess-level-modifieroptstructstruct-namegeneric-
function-signature → parameter-clausethrowsoptfunction-resultopt parameter-clauseopttype-inheritance-clauseoptgeneric-where-clauseoptstruct-body
function-signature → parameter-clauserethrowsfunction-resultopt struct-name → identifier
function-result → ->attributesopttype struct-body → {struct-membersopt}
function-body → code-block struct-members → struct-memberstruct-membersopt
parameter-clause → () (parameter-list) struct-member → declaration compiler-control-statement
parameter-list → parameter parameter,parameter-list
G R A M M A R O F A C L A S S D E C L A R AT I O N
parameter → external-parameter-nameoptlocal-parameter-nametype-
annotationdefault-argument-clauseopt class-declaration → attributesoptaccess-level-modifieroptfinaloptclassclass-
parameter → external-parameter-nameoptlocal-parameter-nametype-annotation namegeneric-parameter-clauseopttype-inheritance-clauseoptgeneric-where-
parameter → external-parameter-nameoptlocal-parameter-nametype-annotation... clauseoptclass-body
external-parameter-name → identifier
local-parameter-name → identifier

251
class-declaration → attributesoptfinalaccess-level-modifieroptclassclass- initializer-declaration → initializer-headgeneric-parameter-clauseoptparameter-
namegeneric-parameter-clauseopttype-inheritance-clauseoptgeneric-where- clausethrowsoptgeneric-where-clauseoptinitializer-body
clauseoptclass-body initializer-declaration → initializer-headgeneric-parameter-clauseoptparameter-
class-name → identifier clauserethrowsgeneric-where-clauseoptinitializer-body
class-body → {class-membersopt} initializer-head → attributesoptdeclaration-modifiersoptinit
class-members → class-memberclass-membersopt initializer-head → attributesoptdeclaration-modifiersoptinit?
class-member → declaration compiler-control-statement initializer-head → attributesoptdeclaration-modifiersoptinit!
initializer-body → code-block
G R A M M A R O F A P R O T O C O L D E C L A R AT I O N
G R A M M A R O F A D E I N I T I A L I Z E R D E C L A R AT I O N
protocol-declaration → attributesoptaccess-level-modifieroptprotocolprotocol-
nametype-inheritance-clauseoptprotocol-body deinitializer-declaration → attributesoptdeinitcode-block
protocol-name → identifier
G R A M M A R O F A N E X T E N S I O N D E C L A R AT I O N
protocol-body → {protocol-membersopt}
protocol-members → protocol-memberprotocol-membersopt extension-declaration → attributesoptaccess-level-modifieroptextensiontype-
protocol-member → protocol-member-declaration compiler-control-statement identifiertype-inheritance-clauseoptextension-body
protocol-member-declaration → protocol-property-declaration extension-declaration → attributesoptaccess-level-modifieroptextensiontype-
protocol-member-declaration → protocol-method-declaration identifiergeneric-where-clauseextension-body
protocol-member-declaration → protocol-initializer-declaration extension-body → {extension-membersopt}
protocol-member-declaration → protocol-subscript-declaration extension-members → extension-memberextension-membersopt
protocol-member-declaration → protocol-associated-type-declaration extension-member → declaration compiler-control-statement
protocol-member-declaration → typealias-declaration
G R A M M A R O F A S U B S C R I P T D E C L A R AT I O N
G R A M M A R O F A P R O T O C O L P R O P E R T Y D E C L A R AT I O N
subscript-declaration → subscript-headsubscript-resultcode-block
protocol-property-declaration → variable-declaration-headvariable-nametype- subscript-declaration → subscript-headsubscript-resultgetter-setter-block
annotationgetter-setter-keyword-block subscript-declaration → subscript-headsubscript-resultgetter-setter-keyword-block
subscript-head → attributesoptdeclaration-modifiersoptsubscriptparameter-clause
G R A M M A R O F A P R O T O C O L M E T H O D D E C L A R AT I O N
subscript-result → ->attributesopttype
protocol-method-declaration → function-headfunction-namegeneric-parameter-
G R A M M A R O F A N O P E R AT O R D E C L A R AT I O N
clauseoptfunction-signaturegeneric-where-clauseopt
operator-declaration → prefix-operator-declaration postfix-operator-declaration
G R A M M A R O F A P R O T O C O L I N I T I A L I Z E R D E C L A R AT I O N
infix-operator-declaration
protocol-initializer-declaration → initializer-headgeneric-parameter- prefix-operator-declaration → prefixoperatoroperator
clauseoptparameter-clausethrowsoptgeneric-where-clauseopt postfix-operator-declaration → postfixoperatoroperator
protocol-initializer-declaration → initializer-headgeneric-parameter- infix-operator-declaration → infixoperatoroperatorinfix-operator-groupopt
clauseoptparameter-clauserethrowsgeneric-where-clauseopt infix-operator-group → :precedence-group-name
G R A M M A R O F A P R O T O C O L S U B S C R I P T D E C L A R AT I O N G R A M M A R O F A P R E C E D E N C E G R O U P D E C L A R AT I O N

protocol-subscript-declaration → subscript-headsubscript-resultgetter-setter- precedence-group-declaration → precedencegroupprecedence-group-

keyword-block name{precedence-group-attributesopt}
precedence-group-attributes → precedence-group-attributeprecedence-group-
G R A M M A R O F A P R O T O C O L A S S O C I AT E D T Y P E D E C L A R AT I O N
attributesopt
protocol-associated-type-declaration → attributesoptaccess-level- precedence-group-attribute → precedence-group-relation
modifieroptassociatedtypetypealias-nametype-inheritance-clauseopttypealias- precedence-group-attribute → precedence-group-assignment
assignmentopt precedence-group-attribute → precedence-group-associativity
precedence-group-relation → higherThan:precedence-group-names
G R A M M A R O F A N I N I T I A L I Z E R D E C L A R AT I O N
precedence-group-relation → lowerThan:precedence-group-names

252
 precedence-group-assignment → assignment:boolean-literal pattern → type-casting-pattern
precedence-group-associativity → associativity:left pattern → expression-pattern
precedence-group-associativity → associativity:right
G R A M M A R O F A W I L D C A R D PAT T E R N
precedence-group-associativity → associativity:none
precedence-group-names → precedence-group-name precedence-group- wildcard-pattern → _
name,precedence-group-names
G R A M M A R O F A N I D E N T I F I E R PAT T E R N
precedence-group-name → identifier
identifier-pattern → identifier
G R A M M A R O F A D E C L A R AT I O N M O D I F I E R
G R A M M A R O F A VA L U E - B I N D I N G PAT T E R N
declaration-modifier → class convenience dynamic final infix lazy optional
override postfix prefix required static unowned unowned(safe) value-binding-pattern → varpattern letpattern
unowned(unsafe) weak
G R A M M A R O F A T U P L E PAT T E R N
declaration-modifier → access-level-modifier
declaration-modifier → mutation-modifier tuple-pattern → (tuple-pattern-element-listopt)
declaration-modifiers → declaration-modifierdeclaration-modifiersopt tuple-pattern-element-list → tuple-pattern-element tuple-pattern-element,tuple-
access-level-modifier → private private(set) pattern-element-list
access-level-modifier → fileprivate fileprivate(set) tuple-pattern-element → pattern identifier:pattern
access-level-modifier → internal internal(set)
G R A M M A R O F A N E N U M E R AT I O N C A S E PAT T E R N
access-level-modifier → public public(set)
access-level-modifier → open open(set) enum-case-pattern → type-identifieropt.enum-case-nametuple-patternopt
mutation-modifier → mutating nonmutating
G R A M M A R O F A N O P T I O N A L PAT T E R N

Attributes optional-pattern → identifier-pattern?

G R A M M A R O F A N AT T R I B U T E G R A M M A R O F A T Y P E C A S T I N G PAT T E R N

attribute → @attribute-nameattribute-argument-clauseopt type-casting-pattern → is-pattern as-pattern

attribute-name → identifier is-pattern → istype
attribute-argument-clause → (balanced-tokensopt) as-pattern → patternastype
attributes → attributeattributesopt G R A M M A R O F A N E X P R E S S I O N PAT T E R N
balanced-tokens → balanced-tokenbalanced-tokensopt
balanced-token → (balanced-tokensopt) expression-pattern → expression
balanced-token → [balanced-tokensopt]
balanced-token → {balanced-tokensopt} Generic Parameters and Arguments
balanced-token → Any identifier, keyword, literal, or operator
G R A M M A R O F A G E N E R I C PA R A M E T E R C L A U S E
balanced-token → Any punctuation except (, ), [, ], {, or }
generic-parameter-clause → <generic-parameter-list>
Patterns generic-parameter-list → generic-parameter generic-parameter,generic-
parameter-list
G R A M M A R O F A PAT T E R N generic-parameter → type-name
pattern → wildcard-patterntype-annotationopt generic-parameter → type-name:type-identifier
pattern → identifier-patterntype-annotationopt generic-parameter → type-name:protocol-composition-type
pattern → value-binding-pattern generic-where-clause → whererequirement-list
pattern → tuple-patterntype-annotationopt requirement-list → requirement requirement,requirement-list
pattern → enum-case-pattern requirement → conformance-requirement same-type-requirement
pattern → optional-pattern conformance-requirement → type-identifier:type-identifier

253
conformance-requirement → type-identifier:protocol-composition-type
same-type-requirement → type-identifier==type
GRAMMAR OF A GENERIC ARGUMENT CLAUSE

generic-argument-clause → <generic-argument-list>
generic-argument-list → generic-argument generic-argument,generic-argument-list
generic-argument → type

254
 Revision History
Revision History

255
 Section 1

Document Revision History

Document Revision History
This table describes the changes to The Swift Programming Language.

Date Notes

Updated for Swift 3.0.1.

Updated the discussion of weak and unowned references in the
Automatic Reference Counting chapter.
Added information about the unowned, unowned(safe), and
201
unowned(unsafe) declaration modifiers in the Declaration
6-1
Modifiers section.
0-2
7 Added a note to the Type Casting for Any and AnyObject section
about using an optional value when a value of type Any is
expected.
Updated the Expressions chapter to separate the discussion of
parenthesized expressions and tuple expressions.

256
 Trademarks
Copyright and Notices APPLE MAKES NO WARRANTY OR REPRESENTATION, EITHER EXPRESS OR
IMPLIED, WITH RESPECT TO THIS DOCUMENT, ITS QUALITY, ACCURACY,
MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. AS A RESULT,
Apple Inc.
THIS DOCUMENT IS PROVIDED “AS IS,” AND YOU, THE READER, ARE
Copyright © 2016 Apple Inc.
ASSUMING THE ENTIRE RISK AS TO ITS QUALITY AND ACCURACY.
All rights reserved.

IN NO EVENT WILL APPLE BE LIABLE FOR DIRECT, INDIRECT, SPECIAL,

INCIDENTAL, OR CONSEQUENTIAL DAMAGES RESULTING FROM ANY DEFECT,
No part of this publication may be reproduced, stored in a retrieval system, or
ERROR OR INACCURACY IN THIS DOCUMENT, even if advised of the possibility of
transmitted, in any form or by any means, mechanical, electronic, photocopying,
such damages.
recording, or otherwise, without prior written permission of Apple Inc., with the
following exceptions: Any person is hereby authorized to store documentation on a
single computer or device for personal use only and to print copies of documentation Some jurisdictions do not allow the exclusion of implied warranties or liability, so the
for personal use provided that the documentation contains Apple’s copyright notice. above exclusion may not apply to you.

No licenses, express or implied, are granted with respect to any of the technology
described in this document. Apple retains all intellectual property rights associated
with the technology described in this document. This document is intended to assist
application developers to develop applications only for Apple-branded products.

Apple Inc.
1 Infinite Loop
Cupertino, CA 95014
408-996-1010

Apple, the Apple logo, Bonjour, Cocoa, Cocoa Touch, Logic, Mac, Numbers,
Objective-C, OS X, Retina, Sand, Shake, and Xcode are trademarks of Apple Inc.,
registered in the U.S. and other countries.

Swift and tvOS are trademarks of Apple Inc.

IOS is a trademark or registered trademark of Cisco in the U.S. and other countries
and is used under license.

Times is a registered trademark of Heidelberger Druckmaschinen AG, available from

Linotype Library GmbH.

257