Swift原型設計模式

Question

有人可以幫我理解這個Swift代碼嗎？此代碼是根據預定我最近在讀的原型設計模式的實現：

class AbstractCard { 
    var name: String? 
    var mana: Int? 
    var attack: Int? 
    var defense: Int? 
    init(name:String?, mana:Int?, attack:Int?, defense:Int?) { 
     self.name = name 
     self.attack = attack 
     self.defense = defense 
     self.mana = mana 
    } 
    func clone() -> AbstractCard { 
     return AbstractCard(name: self.name, mana: self.mana, attack: 
      self.attack, defense: self.defense) 
    } 
} 
class Card: AbstractCard { 
    var someNumber: Int 
    override init(name:String?, mana:Int?, attack:Int?, defense:Int?){ 
     someNumber = 2 
     super.init(name: name,mana: mana,attack: attack,defense: 
      defense) 
    } 
} 
// Simulate our client 
// This is the card that we will copy 
let raidLeader = Card(name: "Raid Leader", mana: 3, attack: 2, defense: 2) 
// Now we use our faceless Manipulator card to clone the 

let facelessManipulator = raidLeader.clone() 
print("\(facelessManipulator.name, facelessManipulator.mana, facelessManipulator.attack, facelessManipulator.defense)") 
print(facelessManipulator.dynamicType) // prints "AbstractCard" 

這是什麼模式的點，如果克隆的對象的dynamicType仍然AbstractCard，而不是卡。你甚至不能訪問特定於卡的變量。我試圖施放此對象卡，但我得到的錯誤"Execution was interrupted reason: signal SIGABRT"

Answer 1

讓我們明確了原型模式第一：

原型模式通過複製現有的對象，被稱爲原型創建新的對象。

首先有很多方法來實現Prototype模式，我認爲本書的作者試圖以最合適的方式解釋你模式的用戶案例。

因此，作爲擺在評論別人的clone方法應該是覆蓋在子類，允許對象的克隆正確，如以下方式：

class Card: AbstractCard { 

    var someNumber: Int 

    override init(name:String?, mana:Int?, attack:Int?, defense:Int?){ 
     someNumber = 2 
     super.init(name: name,mana: mana,attack: attack,defense: 
     defense) 
    } 

    override func clone() -> AbstractCard { 
     return Card(name: self.name, mana: self.mana, attack: 
     self.attack, defense: self.defense) 
    } 
} 

我必須說，我曾作爲本書的技術評論員，我認爲作者用更好的方式表達了他可以使用這種模式。

如果您認真思考，每次在Swift（Arrays，Int，Bool等）中使用值類型時都會應用此模式，因爲每次將實例的值分配給另一個實例時，都會創建一個新副本相同的值。問題在於引用值（類），每次將實例分配給新實例時，它們都會共享其值和修改。

實施該模式的另一種方式是使用協議，該協議聲明瞭提供對象的功能副本的方法，並且在我看來是最常用的方法。

例如：

class Card: NSObject, NSCopying { 

    var name: String? 
    var mana: Int? 
    var attack: Int? 
    var defense: Int? 

    init(name:String?, mana:Int?, attack:Int?, defense:Int?) { 
     self.name = name 
     self.attack = attack 
     self.defense = defense 
     self.mana = mana 
    } 

    func copyWithZone(zone: NSZone) -> AnyObject { 
     return Card(name: self.name, mana: self.mana, attack: self.attack, defense: self.defense) 
    } 
} 

// Simulate our client 
// This is the card that we will copy 
let raidLeader = Card(name: "Raid Leader", mana: 3, attack: 2, defense: 2) 
// Now we use our faceless Manipulator card to clone the 

let facelessManipulator = raidLeader.copy() as Card 
raidLeader.attack = 5 

print("\(facelessManipulator.name, facelessManipulator.mana, facelessManipulator.attack, facelessManipulator.defense)") 
print("\(raidLeader.name, raidLeader.mana, raidLeader.attack, raidLeader.defense)") 

您可以瞭解這個問題，更多是這樣的：

• Implementing NSCopying in Swift with subclasses

我希望這可以幫助您。