How to include completion handlers for multiple functions in swift? - swift

Question

Consider this code:
func test() {
A()
B()
C()
D()
E()
}
Each function here have their own set of actions like calling API's, parsing them, writing results in file, uploading to servers, etc etc.
I want to run this functions one by one. I read about completion handlers. My problem with completion handlers are:
All examples given to understand completion handlers was having just two methods
I don't want to place this functions in some other function. I want all function calls (A to E) inside Test() function alone
Can someone help on this?

Answer 1

This is easy to do, you just need to add a closure argument to call on completion. For example:
func a(completion: (() -> Void)) {
// When all async operations are complete:
completion()
}
func b(completion: (() -> Void)) {
// When all async operations are complete:
completion()
}
func c(completion: (() -> Void)) {
// When all async operations are complete:
completion()
}
func d(completion: (() -> Void)) {
// When all async operations are complete:
completion()
}
func e(completion: (() -> Void)) {
// When all async operations are complete:
completion()
}
func test() {
a {
b {
c {
d {
e {
// All have now completed.
}
}
}
}
}
}
As you can see, this looks horrible. One problem with multiple async operations, non concurrently is that you end up with this horrible nesting.
Solutions to this do exist, I personally recommend PromiseKit. It encapsulates blocks in easy chaining methods, which is far cleaner.

Answer 2

Take a look at reactive programming and observable chains. RxSwift will be an excellent library for you. Reactive programming is very popular right now and was created exactly to solve problems like yours. It enables you to easily create a process (like an assembly line) that transforms inputs into your desired outputs. For example, your code in Rx would look like:
A()
.flatMap { resultsFromA in
B(resultsFromA)
}
.flatMap { resultsFromB in
C(resultsFromB)
}
.flatMap { resultsFromC in
D(resultsFromC)
}
.flatMap { resultsFromD in
E(resultsFromD)
}
.subscribe(onNext: { finalResult in
//result from the process E, subscribe starts the whole 'assembly line'
//without subscribing nothing will happen, so you can easily create observable chains and reuse them throughout your app
})
This code sample would create a process where you'd transform results from your initial API call (A), into parsed results (B), then write parsed results to a file (C) etc. Those processes can be synchronous or asynchronous. Reactive programming can be understood as Observable pattern on steroids.