A guide to our Swift style and conventions.
This is an attempt to encourage patterns that accomplish the following goals (in rough priority order):
- Increased rigor, and decreased likelihood of programmer error
- Increased clarity of intent
- Reduced verbosity
- Fewer debates about aesthetics
If you have suggestions, please see our contribution guidelines, then open a pull request. ⚡
- Tabs, not spaces.
- End files with a newline.
- Make liberal use of vertical whitespace to divide code into logical chunks.
- Don’t leave trailing whitespace.
- Not even leading indentation on blank lines.
When possible, omit the get
keyword on read-only computed properties and
read-only subscripts.
So, write these:
var myGreatProperty: Int {
return 4
}
subscript(index: Int) -> T {
return objects[index]
}
… not these:
var myGreatProperty: Int {
get {
return 4
}
}
subscript(index: Int) -> T {
get {
return objects[index]
}
}
Rationale: The intent and meaning of the first version is clear, and results in less code.
Top-level functions, types, and variables should always have explicit access control specifiers:
public var whoopsGlobalState: Int
internal struct TheFez {}
private func doTheThings(things: [Thing]) {}
However, definitions within those can leave access control implicit, where appropriate:
internal struct TheFez {
var owner: Person = Joshaber()
}
Rationale: It's rarely appropriate for top-level definitions to be specifically internal
, and being explicit ensures that careful thought goes into that decision. Within a definition, reusing the same access control specifier is just duplicative, and the default is usually reasonable.
When specifying the type of an identifier, always put the colon immediately after the identifier, followed by a space and then the type name.
class SmallBatchSustainableFairtrade: Coffee { ... }
let timeToCoffee: NSTimeInterval = 2
func makeCoffee(type: CoffeeType) -> Coffee { ... }
Rationale: The type specifier is saying something about the identifier so it should be positioned with it.
When accessing properties or methods on self
, leave the reference to self
implicit by default:
private class History {
var events: [Event]
func rewrite() {
events = []
}
}
Only include the explicit keyword when required by the language—for example, in a closure, or when parameter names conflict:
extension History {
init(events: [Event]) {
self.events = events
}
var whenVictorious: () -> () {
return {
self.rewrite()
}
}
}
Rationale: This makes the capturing semantics of self
stand out more in closures, and avoids verbosity elsewhere.
Unless you require functionality that can only be provided by a class (like identity or deinitializers), implement a struct instead.
Note that inheritance is (by itself) usually not a good reason to use classes, because polymorphism can be provided by protocols, and implementation reuse can be provided through composition.
For example, this class hierarchy:
class Vehicle {
let numberOfWheels: Int
init(numberOfWheels: Int) {
self.numberOfWheels = numberOfWheels;
}
func maximumTotalTirePressure(pressurePerWheel: Float) -> Float {
return pressurePerWheel * numberOfWheels
}
}
class Bicycle: Vehicle {
init() {
super.init(numberOfWheels: 2)
}
}
class Car: Vehicle {
init() {
super.init(numberOfWheels: 4)
}
}
could be refactored into these definitions:
protocol Vehicle {
var numberOfWheels: Int { get }
}
func maximumTotalTirePressure(vehicle: Vehicle, pressurePerWheel: Float) -> Float {
return pressurePerWheel * vehicle.numberOfWheels
}
struct Bicycle: Vehicle {
let numberOfWheels = 2
}
struct Car: Vehicle {
let numberOfWheels = 4
}
Rationale: Value types are simpler, easier to reason about, and behave as expected with the let
keyword.
Classes should start as final
, and only be changed to allow subclassing if a valid need for inheritance has been identified. Even in that case, as many definitions as possible within the class should be final
as well, following the same rules.
Rationale: Composition is usually preferable to inheritance, and opting in to inheritance hopefully means that more thought will be put into the decision.
Methods of parameterized types can omit type parameters on the receiving type when they’re identical to the receiver’s. For example:
struct Composite<T> {
…
func compose(other: Composite<T>) -> Composite<T> {
return Composite<T>(self, other)
}
}
could be rendered as:
struct Composite<T> {
…
func compose(other: Composite) -> Composite {
return Composite(self, other)
}
}
Rationale: Omitting redundant type parameters clarifies the intent, and makes it obvious by contrast when the returned type takes different type parameters.