This is a creation in Article, where the information may have evolved or changed.
The simple factory is mainly used to solve the problem of object "creation". The following example is taken from the first chapter of "Big Talk design mode" to implement an extensible "calculator". When adding new functionality, there is no need to change the algorithm that was already implemented. Because it is a simple factory, we still need to modify the factory class accordingly.
1. First, we define a computational interface
package calctypeinterface { SetData(data ...interface{}) float64}
2. Next, we implement two subclasses of this class, namely addition and subtraction
addition, which is added by two numbers:)
PackageCalcImport "FMT"typeAddstruct{NUM1float64Num2float64}funcNewadd () *add {instance: =New(ADD)returnInstancefunc(A *add) SetData (Data ...Interface{}) {if Len(data)! =2{FMT. Println ("error,need-parameters")return}if_, OK: = Data[0]. (float64);!ok {fmt. Println ("error,need float64 parameters")return}if_, OK: = Data[1]. (float64);!ok {fmt. Println ("error,need float64 parameters")return} A.NUM1 = Data[0]. (float64) a.num2 = Data[1]. (float64)}func(A Add) Calcoperate ()float64{returnA.num1 + a.num2}
Subtraction, is to subtract two numbers:), I feel so cold ...
PackageCalcImport "FMT"typeSubtractionstruct{NUM1float64Num2float64}funcNewsubtraction () *subtraction {instance: =New(subtraction)returnInstancefunc(A *subtraction) SetData (Data ...Interface{}) {if Len(data)! =2{FMT. Println ("error,need-parameters")return}if_, OK: = Data[0]. (float64);!ok {fmt. Println ("error,need float64 parameters")return}if_, OK: = Data[1]. (float64);!ok {fmt. Println ("error,need float64 parameters")return} A.NUM1 = Data[0]. (float64) a.num2 = Data[1]. (float64)}func(A subtraction) Calcoperate ()float64{returnA.NUM1-A.NUM2}
3. It's time to get it done, define a simple factory to instantiate these two classes
PackageCalctypeCalcfactorystruct{}funcNewcalcfactory () *calcfactory {instance: =New(calcfactory)returnInstancefunc(f calcfactory) Createoperate (OpTypestring) Calcsuper {varOp calcsuperSwitchOpType { Case "+": op = Newadd () Case "-": op = newsubtraction ()default:Panic("Error! Dont have this operate ") }returnOp
In this simple factory, we only pass in the corresponding operation method, such as "+", "-", to create the relevant arithmetic strategy. It will return an instance of an operation interface, and when we get this instance, we can invoke the method inside to perform the operation.
4. Testing
Simple Factory mode Project main. GoPackage Mainimport (."Calc" "FMT") Func Main () {factory: = Newcalcfactory () op: = Factory. Createoperate("+") op. SetData(1.0,2.0) FMT. Println(OP. Calcoperate()) op = Factory. Createoperate("-") op. SetData(1.0,2.0) FMT. Println(OP. Calcoperate())/* Output: 3-1 */}
Now, our simple factory model is done, and when we need to add multiplication, we do not avoid modifying or compiling the completed addition and subtraction operations. We just need to add the corresponding structure of the multiplication, and with a new switch branch in the factory, we can complete the multiplication.
Because the go language does not allow the existence of sub-abstract classes, which allows me to implement a very strange function setdata (data ... interface{}). It can pass in multiple arbitrary objects to be used as the data we need in our structure. This is just an example of what the simple factory has to do, and you can refer to the 6th chapter in refactoring and Patterns: knowledge of creation . It's a very comprehensive story.
Copyright NOTICE: This article for Bo Master original article, without Bo Master permission not reproduced.