When to use CodingKeys in Decodable(Swift) - swift

Question

Let's say I want to decode a Person struct as follows.
struct Person: Decodable {
let firstName: String
let lastName: String
let age: Int: String
}
I understand that the data can be decoded only with above. Therefore if I'm not changing the properties to a custom name if there no difference between the above and below implementation?
Further is there other cases where you want to use CodingKeys? I'm confused when they are necessary other than for renaming purposes.
struct Person: Decodable {
let firstName: String
let lastName: String
let age: Int: String
}
enum CodingKeys: String, CodingKey {
case firstName
case lastName
case age
}

Answer 1

First of all there is a make-or-break rule for using CodingKeys:
You can omit CodingKeys completely if the JSON – or whatever Codable conforming format – keys match exactly the corresponding properties (like in your example) or the conversion is covered by an appropriate keyDecodingStrategy.
Otherwise you have to specify all CodingKeys you need to be decoded (see also reason #3 below).
There are three major reasons to use CodingKeys:
A Swift variable/property name must not start with a number. If a key does start with a number you have to specify a compatible CodingKey to be able to decode the key at all.
You want to use a different property name.
You want to exclude keys from being decoded for example an id property which is not in the JSON and is initialized with an UUID constant.
And CodingKeys are mandatory if you implement init(from decoder to decode a keyed container.

Answer 2

You can use CodingKeys in different ways for example, when you know that at least one of the name of values that you are expecting in your JSON is actually different from your "let or var" name.
Example:
struct Person: Decodable {
let firstName: String
let lastName: String
let age: Int: String
}
enum CodingKeys: String, CodingKey {
case firstName = "first_name"
case lastName
case age
}
Other case is when you are using class inheritance.
In conclusion, if you are absolutely sure that you are using the same variable name as your encoding key(JSON), you can omit it (but if you want to put it, it doesn't matter), but if there's a difference, maybe a change of your codingKeys like an uppercase or using different words, you should use the enum to map the correct key with the variable name.

Answer 3

CodingKeys can be extremely helpful if you have a JSON with arbitrary number of coding keys (also called dynamic keys). Here is an example.
import UIKit
// Consider JSON with infinite number of keys: "S001", "S002" and so on
let jsonData = """
{
"S001": {
"firstName": "Tony",
"lastName": "Stark"
},
"S002": {
"firstName": "Peter",
"lastName": "Parker"
},
"S003": {
"firstName": "Bruce",
"lastName": "Wayne"
}
}
""".data(using: .utf8)!
struct Student: Decodable {
let firstName: String
let lastName: String
}
struct DecodedArray: Decodable {
var array: [Student]
// Define DynamicCodingKeys type needed for creating
// decoding container from JSONDecoder
private struct DynamicCodingKeys: CodingKey {
// Use for string-keyed dictionary
var stringValue: String
init?(stringValue: String) {
self.stringValue = stringValue
}
// Use for integer-keyed dictionary
var intValue: Int?
init?(intValue: Int) {
// We are not using this, thus just return nil
return nil
}
}
init(from decoder: Decoder) throws {
// 1
// Create a decoding container using DynamicCodingKeys
// The container will contain all the JSON first level key
let container = try decoder.container(keyedBy: DynamicCodingKeys.self)
var tempArray = [Student]()
// 2
// Loop through each key (student ID) in container
for key in container.allKeys {
// Decode Student using key & keep decoded Student object in tempArray
let decodedObject = try container.decode(Student.self, forKey: DynamicCodingKeys(stringValue: key.stringValue)!)
tempArray.append(decodedObject)
}
// 3
// Finish decoding all Student objects. Thus assign tempArray to array.
array = tempArray
}
}
let decodedResult = try! JSONDecoder().decode(DecodedArray.self, from: jsonData)

Answer 4

Therefore if I'm not changing the properties to a custom name if there no difference between the above and below implementation?
Yes, but there's a bit of misunderstanding here. The two implementations you have are literally identical because in the second one the CodingKeys enum would never be used. To be used, the enum must be nested within the Decodable conforming type (Person in this case):
struct Person: Decodable {
let firstName: String
let lastName: String
let age: Int
enum CodingKeys: String, CodingKey {
case firstName
case lastName
case age
}
}
There is in practice no difference between this implementation and the ones you provided.
Further is there other cases where you want to use CodingKeys?
CodingKeys are not used solely by Decodable, they are also used by Encodable. When using Encodable, a reason to use CodingKeys is to specify only a subset of the instances fields should be serialized.