I want to be able to pass a class (not an initialized object) of a certain protocol type to a method, then call the class functions of that class in the method. Code below.
I am using Swift and have an protocol defined like this
//Protocol for any object to be used with an FAUAPIConnection
protocol FAUAPIModel{
//Used to parse the object from a given dictionary to an object
class func parseFromJSON(JSON:AnyObject) -> Self
//Required default init
init()
}
What I would like to do is have a method like this
func getSomeParsingDone<T:FAUAPIModel>(model:T.Type? = nil, getPath:path, callingObj:CallingClass) -> Void
{
//GetIt is inconsequential, just logic to get an object from a certain path
var returnObj:AnyObject = GetIt.get(path)
if(model != nil){
returnObj = model!.parseFromJSON() <<<<<< Type 'T' does not conform to protocol 'AnyObject'
}
callingObj.done(returnObj)
}
Object that implements the protocol
import Foundation
class MyObj: FAUAPIModel{
var neededVal:String
var nonneededVal:String
required convenience init(){
self.init(neededVal:"VALUE")
}
init(neededVal:String, nonneededVal:String = ""){
self.neededVal = neededVal
self.nonneededVal = nonneededVal
}
class func parseFromJSON(JSON:AnyObject) -> WGMPart
{
return WGMPart() <<<<<<<< Method 'parseFromJSON' in non-final class 'WGMPart' must return 'Self' to conform to protocol 'FAUAPIModel'
}
}
However, I keep getting two errors. I have indicated these above with '<<<<<<<<<<<<'
compile error.
Lots of little things to consider here, but let's get to the heart of your question. The signature you want looks like this:
func getSomeParsingDone<T:FAUAPIModel>(model:T.Type, path:String) -> T?
I'm making the return optional beause there are a lot of things that could fail here, and you really shouldn't turn all of those into crashes.
I'd recommend your protocol look like this:
protocol FAUAPIModel {
class func parseFromJSON(JSON:AnyObject) -> Self
}
That way, you're promising that your return your own class, not just anything that is parseable. That does tend to mean that you need to make your classes final. If you don't want them to be final, you'll need to promise some init method in order to construct it. See Protocol func returning Self for more details on how to deal with that if you need it.
So putting it together, it might look something like this in practice:
protocol FAUAPIModel {
class func parseFromJSON(JSON:AnyObject) -> Self
}
func createObjectOfClass<T: FAUAPIModel>(model: T.Type, path: String) -> T? {
if let json: AnyObject = GetJSON(path) {
return model.parseFromJSON(json)
}
return nil
}
// Bogus JSON reader
func GetJSON(path: String) -> AnyObject? {
let json: AnyObject? = NSJSONSerialization.JSONObjectWithData(path.dataUsingEncoding(NSUTF8StringEncoding, allowLossyConversion: true)!, options: NSJSONReadingOptions(0), error: nil)
return json
}
// Bogus model class that returns trivial version of itself
final class Something: FAUAPIModel {
class func parseFromJSON(JSON:AnyObject) -> Something {
return Something()
}
}
// Using it
let something = createObjectOfClass(Something.self, "/path/to/file")
I just want to note that the answer to your exact question would be to declare your function like this:
func getSomeParsingDone(model:FAUAPIModel.Type? = nil, getPath:path) -> FAUAPIModel
Related
I'm trying to get a container that implements a set of protocols that i pass as parameter to a function on the original container.
struct Container {
let someProperty: String
let otherProperty: String
}
// Subcontainers
protocol Base {}
protocol SomePropertyContainer: Base {
var someProperty: String { get }
}
protocol OtherPropertyContainer: Base {
var otherProperty: String { get }
}
extension Container: SomePropertyContainer, OtherPropertyContainer {}
// Sub Container Provisioning Protocol
protocol SubContainerProviderProtocol {
func subContainer<T: Base>(protos: T.Type) -> T?
}
extension Container: SubContainerProviderProtocol {
func subContainer <T: Base>(protos: T.Type) -> T? {
return self as? T
}
}
// Example container
let subContainerProvider: SubContainerProviderProtocol = Container(someProperty: "Why does this not work!", otherProperty: "Seriously.")
Getting this up and running would allow me to inject the ContainerProviderProtocol into consumers while giving them the possibility to specify themselves which SubContainer they actually want.
E.g. a class that would be interested in only the someProperty could look like this
// Example Container Provider consumer
class SomeClass {
let subContainerProvider: SubContainerProviderProtocol
init(subContainerProvider: SubContainerProviderProtocol) {
self.subContainerProvider = subContainerProvider
}
func printSomeProperty() {
let someProperty = subContainerProvider
.subContainer(protos: SomePropertyContainer.self)?
.someProperty
print(someProperty)
}
}
// Example call
let someClass = SomeClass(subContainerProvider: subContainerProvider)
someClass.printSomeProperty() // "Why does this not work!"
This solution would be incredible for dependency injection & testability.
However the restriction T: Base is causing the compiler error
In argument type 'SomePropertyContainer.Protocol', 'SomePropertyContainer' does not conform to expected type 'Base'
Not specifying conformance to Base will compile, but would also allow to pass any type as T.
I've tried with associated types within an additional protocol etc, however have not figured it out. And while this issue is incredibly fun, I'm running out of ideas.
Possibly related to (but not exactly same) https://bugs.swift.org/browse/SR-55
Here's the problem: at some point you have to start working with actual types, and not just protocols. Your line:
func container<T: Base>(protos: T.Type) -> T?
is telling the compiler that you're going to give this function a type, generically T, that conforms to the protocol Base, not another protocol. You need something like this:
class SPC: SomePropertyContainer {
var someProperty: String = ""
}
class SomeClass {
let containerProvider: ContainerProviderProtocol
init(containerProvider: ContainerProviderProtocol) {
self.containerProvider = containerProvider
}
func printSomeProperty() {
let someProperty = containerProvider
.container(protos: SPC.self)?
.someProperty
print(someProperty)
}
}
SPC is a type that conforms to the SomePropertyContainer protocol, which itself conforms to the Base protocol, so this is what your code is expecting.
I have a scenario where I need to call a method with an optional closure argument where the closure receives a generic argument. Here's my simplified code:
class FooModel
{
}
class FooSubClass1 : FooModel
{
}
class FooSubClass2 : FooModel
{
}
class Client
{
func httpGet<T:FooModel>(closure:((T) -> Void)? = nil)
{
// Doing some network request stuff here and call onHTTPRequestComplete() when done!
onHTTPRequestComplete(data: result, closure: closure)
}
func onHTTPRequestComplete<T:FooModel>(data:[String:Any], closure:((T) -> Void)? = nil)
{
if let c = closure
{
switch(T)
{
case is FooSubClass1:
var foo = FooSubClass1()
// Process data into new FooSubClass1 object here!
c(foo)
case is FooSubClass2:
var foo = FooSubClass2()
// Process data into new FooSubClass2 object here!
c(foo)
}
}
}
}
class App
{
func someFunc()
{
client.httpGet()
{
response in
print("\(response)")
}
}
}
What I'm trying to do here:
I have a data model with a super class and several subclasses. In Client I do http requests to retrieve data and want to populate model objects with the data retrieved. The calling method in App should be able to provide a closure that is called after the asynchronous network request is done and in which I get back the correct data model object.
There are several issues with my code:
switch(T): Error: expected member name or constructor call after type name
c(foo): Error: '(FooSubClass1) -> Void' is not convertible to '(T) -> Void'
Not sure my closure definition makes sense for what I'm trying to do.
Updated code:
protocol FooModel
{
init()
}
class FooModelBase : FooModel
{
}
class FooSubClass1 : FooModelBase
{
}
class FooSubClass2 : FooModelBase
{
}
class Client
{
func httpGet<T:FooModel>(closure:((T) -> Void)? = nil)
{
// Doing some network request stuff here and call onHTTPRequestComplete() when done!
onHTTPRequestComplete(data: result, closure: closure)
}
func onHTTPRequestComplete<T:FooModel>(data:[String:Any], closure:((T) -> Void)? = nil)
{
if let c = closure
{
switch T.self
{
case is FooSubClass1.Type:
var foo = FooSubClass1()
// Assign retrieved values to model!
closure(foo)
case is FooSubClass2.Type:
var foo = FooSubClass2()
// Assign retrieved values to model!
closure(foo)
default:
break
}
}
}
}
class App
{
func someFunc()
{
client.httpGet()
{
response in
print("\(response)") // Should output FooSubClass1 instance with assigned values!
}
}
}
Almost works but I get the following compile error:
'closure(foo): Error: '(#lvalue FooSubClass1) -> Void' is not convertible to '(T) -> Void'
for the switch(T) error of course you get this error. Because you're trying to switch a type but you should be switching a value. For your second error it makes perfect sense because then what would closure be expecting then because T is generic.
For more info on generics in closures have a look at this stackoverflow thread
generics as parameters to a closure in swift
All in all what you're trying to do there just isn't possible and at first glance I believe this isn't the correct direction to go in (Although I could be wrong without seeing the whole context). One thing you can do instead maybe is have your closure be ((Any) -> Void)? but then unfortunately the type would be of Any and you would have to do a switch inside of the closure.
The value of T in the closure will be passed to the closure itself. You can't switch on it. But what I think you want is to switch on the type. You could do something like this:
let type = String(reflecting: T.self)
switch (type) {
case "Module.FooSubClass1":
var foo = FooSubClass1()
}
Or you could say:
if FooSubClass1.self == T.self {
var foo = FooSubClass1();
} else if FooSubClass2.self == T.self {
var foo = FooSubClass2();
}
However, this code smells. You seem to want to use generics but then have a hardcoded set of classes inside your generic method, which half-defeats the entire purpose of generics. Why not just define FooModel as
protocol FooModel {
init()
}
Then instead of your switch statement, just create one and pass it to your closure.
func onHTTPRequestComplete<T: FooModel>(data: [String: Any], closure: ((T) -> Void)? = nil) {
guard let closure = closure else { return }
closure(T())
}
Or why not create T outside of onHTTPRequestComplete and pass it?
onHTTPRequestComplete(data: data: model: FooSubClass1()) { model in
// Do something awesome with `model` right here.
}
However, what I'd probably do is this:
func httpGet(callback: (([String: Any]) -> Void)? = nil) {
// Do some network stuff and get the data
callback?(data)
}
Then inside of callback, just create whatever type you want:
httpGet { data in
let foo = FooSubClass1(data: data)
// foo on you
}
I try to call a class method on a generic T: BaseModel where T can be a subclass of BaseModel.
For example Car.
In the case T should be Car, I want my class method to be called on the Car class.
However, It always ends up calling the BaseModel class method instead.
class func parse<T: BaseModel>(json: JSON, context: NSManagedObjectContext) throws -> T? {
return T.classParseMethod(json: json) //This never calls the Car.classParseMethod()
}
where
let carObject = parse(json:json, context:context) as? Car
Any help?
The casting is done after the function call so the generic constraint resolves to T = BaseModel. You want the function to know of the type so it can properly resolve the generic constraint:
func parse<T: BaseModel>(_ str: String) -> T? {
print(T.Type.self) // should print like: Car.Type
return T.parse(str) as? T
}
// Make the desired type known to swift
let car: Car? = parse("My Car Format String")
One solution that seems to work is:
to add
func myFunc<T: BaseModel>(_ type: T.Type,..) -> T? {
type.aClassFunc()
{
If I call the following, it works.
if let obj = myFunc(Car.self, ...) {
// obj will be of type Car
}
It seems really too much just to achieve this but there might be an underlying reason for it.
Is there a point to declaring a static function on a protocol? The client using the protocol has to call the function on a type conforming to the protocol anyway right? That breaks the idea of not having to know the type conforming to the protocol IMO. Is there a way to call the static function on the protocol in a way where I don't have to know the actual type conforming to my protocol?
Nice question. Here is my humble point of view:
Is there a point to declaring a static function on a protocol?
Pretty much the same as having instance methods declared in a protocol.
The client using the protocol has to call the function on a type conforming to the protocol anyway right?
Yes, exactly like instance functions.
That breaks the idea of not having to know the type conforming to the protocol IMO.
Nope. Look at the following code:
protocol Feline {
var name: String { get }
static func createRandomFeline() -> Feline
init()
}
extension Feline {
static func createRandomFeline() -> Feline {
return arc4random_uniform(2) > 0 ? Tiger() : Leopard()
}
}
class Tiger: Feline {
let name = "Tiger"
required init() {}
}
class Leopard: Feline {
let name = "Leopard"
required init() {}
}
let feline: Feline = arc4random_uniform(2) > 0 ? Tiger() : Leopard()
let anotherFeline = feline.dynamicType.createRandomFeline()
I don't know the real type inside the variable feline. I just know that it does conform to Feline. However I am invoking a static protocol method.
Is there a better way to do this?
I see, you would like to call a static method/function declared in a protocol without creating a value that conforms to the protocol.
Something like this:
Feline.createRandomFeline() // DANGER: compiler is not happy now
Honestly I don't know the reason why this is not possible.
yes this is possible:
Swift 3
protocol Thing {
static func genericFunction()
}
//... in another file
var things:[Thing] = []
for thing in things {
type(of: thing).genericFunction()
}
Thank you #appzYourLife for the help! Your answer inspired my answer.
#appzYourLife answered my question. I had an underlying issue I was trying to resolve and the following code resolves my issue, so I'll post this here, maybe it helps someone with my same underlying question:
protocol MyProtocol {
static func aStaticFunc()
}
class SomeClassThatUsesMyProtocolButDoesntConformToIt {
var myProtocolType: MyProtocol.Type
init(protocolType: MyProtocol.Type) {
myProtocolType = protocolType
}
func aFunction() {
myProtocolType.aStaticFunc()
}
}
I created another solution for this case. IMHO this is quite clean and simple.
First, create a protocol for accessing instance type.
protocol TypeAccessible {
func type() -> Self.Type
}
extension TypeAccessible {
func type() -> Self.Type {
return Swift.type(of: self)
}
}
then create your concrete class as here. The point is your protocol should conform to TypeAccessible protocol.
protocol FooProtocol: TypeAccessible {
static func bar()
}
class Foo: FooProtocol {
static func bar() { }
}
On call site use it as
let instance: FooProtocol = Foo()
instance.type().bar()
For further use cases, just make sure your protocols conform to TypeAccessible and that's all.
A little late to the party on this one.
Here's my solution for "adding" static properties/functions/types to a protocol using typealias.
For example:
enum PropertyScope {
case all
case none
}
struct PropertyNotifications {
static var propertyDidChange =
Notification.Name("propertyDidChangeNotification")
}
protocol Property {
typealias Scope = PropertyScope
typealias Notifications = PropertyNotifications
var scope: Scope { get set }
}
Then you can do this anywhere in your code:
func postNotification() {
let scope: Property.Scope = .all
NotificationCenter.post(name: Property.Notifications.propertyDidChange,
object: scope)
}
Using protocols like Java interfaces is rarely a good idea. They are meta types, meant for defining contracts, which is an entirely different kind of thing.
That being said, just for the point of understanding, I find the most simple and effective way for creating the equivalent of a static factory method of a protocol to write a free function.
It should contain the protocol's name, hoping that that will prevent name clashes, and improve discoverability.
In other languages, createP would be a static member of P, named create and be called as P.create(...), which would drastically improve discoverability and guarantee to prevent name clashes.
In swift, though, this is not an option for protocols, so if protocols are for some reason really actually used as a replacement for interfaces, at least including the protocol's name in the function's name is an ugly workaround that's still slightly better than nothing.
P.S. in case the goal is actually to achieve something like an inheritance hierarchy with structs, union style enums are the tool that's meant to serve that purpose :)
protocol P
{
var x: Int { get }
}
func createP() -> P
{
if (todayIsMonday())
{
return A()
}
else
{
return B()
}
}
class A: P
{
var x = 5
}
class B: P
{
var x = 7
}
This isn't an answer so much as it is an extension to the question. Say I have:
#objc public protocol InteractivelyNameable: Nameable {
static func alertViewForNaming(completion:#escaping((_ success: Bool, _ didCancel: Bool, _ error: Error?) -> Void)) -> UIAlertController?
}
And I have a generic view controller that manages various types (generic type is .fetchableObjectType... basically NSFetchResult). I need to check if a specific object type conforms to the protocol, and if so, invoke it.
something like:
// valid swift code
if self.dataSource.fetchableObjectType is InteractivelyNameable {
// not valid swift code
if let alert = (self.dataSource.fetchableObjectType as InteractivelyNameable).alertViewForNaming(....)
}
I had a situation where I need to create same DomainModel object from 2 different response. so this (static method in protocol helped me) approach helped me.
protocol BaseResponseKeyList: CodingKey {
static func getNameKey()->Self
}
enum FirstResponseKeyList: String, BaseResponseKeyList {
case name
func getNameKey()->FirstResponseKeyList {
return .name
}
}
enum SecondResponseKeyList: String, BaseResponseKeyList {
case userName
func getNameKey()->SecondResponseKeyList {
return .userName
}
}
struct MyDomainModel<T:BaseResponseKeyList> : Decodable {
var name:String?
required init(from d:Decoder) {
do {
let container = try d.container(keyedBy:T.self)
name = try container.decode(String.self, forKey:T.getNameKey())
}catch(_) {
print("error")
}
}
}
let myDomainModel = try JSONDecoder().decode(MyDomainModel <FirstResponseKeyList>.self, from: data)
let myDomainModel2 = try JSONDecoder().decode(MyDomainModel <SecondResponseKeyList>.self, from: data2)
I am writing method which takes a type which conforms to a protocol and instantiates an instance of this class. When I build it, the compiler crashes with a segfault. I appreciate that this points to a compiler bug 99% of the time, but I am interested to see if what I'm trying to do is logically correct or am I just throwing absolute nonsense at the compiler and I shouldn't be surprised to see it crash.
Here is my code
protocol CreatableClass {
init()
}
class ExampleClass : CreatableClass {
required init() {
}
}
class ClassCreator {
class func createClass(classType: CreatableClass.Type) -> CreatableClass {
return classType()
}
}
ClassCreator.createClass(ExampleClass.self)
I also tried to rule out passing a Type as a method parameter as being the root of the problem and the following code also crashes the compiler:
protocol CreatableClass {
init()
}
class ExampleClass : CreatableClass {
required init() {
}
}
let classType: CreatableClass.Type = CreatableClass.self
let instance = classType()
So - is this just a straightforward compiler bug and does what I am trying to do seem reasonable, or is there something in my implementation that is wrong?
Edit:
This can be achieved using generics as shown #Antonio below but unfortunately i believe that isn't useful for my application.
The actual non-dumbed down use-case for doing this is something like
protocol CreatableClass {}
protocol AnotherProtocol: class {}
class ClassCreator {
let dictionary: [String : CreatableClass]
func addHandlerForType(type: AnotherProtocol.Type, handler: CreatableClass.Type) {
let className: String = aMethodThatGetsClassNameAsAString(type)
dictionary[className] = handler()
}
required init() {}
}
I usually do that by defining a generic method. Try this:
class func createClass<T: CreatableClass>(classType: T.Type) -> CreatableClass {
return classType()
}
Update
A possible workaround is to pass a closure creating a class instance, rather than passing its type:
class ClassCreator {
class func createClass(instantiator: () -> CreatableClass) -> (CreatableClass, CreatableClass.Type) {
let instance = instantiator()
let classType = instance.dynamicType
return (instance, classType)
}
}
let ret = ClassCreator.createClass { ExampleClass() }
The advantage in this case is that you can store the closure in a dictionary for example, and create more instances on demand by just knowing the key (which is something in 1:1 relationship with the class name).
I used that method in a tiny dependency injection framework I developed months ago, which I realized it works only for #objc-compatible classes only though, making it not usable for my needs...