Now in my code, I have a bunch of repetitive work that looks like this.
type fp = func(arg1 string, arg2 int, arg3 []string) bool func decorator(fp fp, arg1 string, arg2 int, arg3 []string) { // do some stuff prior to running fp res := fp(arg1, arg2, arg3) // do some other stuff depending on res }
My problem is that every time I want to add a new fp type with a different number/type of parameters, I have to redefine this method, so I end up redefining the decorator 4-5 times in my code, and It may take a little more work. I want to replace
with decoratortype FP = func(a ...interface{}) bool func Decorator(fp FP, argDetails interface{}) { // do stuff res := fp(a) // do other stuff }
And let fp assert the type. But I'm not sure if this causes problems/is a bad use case as I've seen a lot about not using the interface/causing performance issues.
For some reasons, generics are not helpful, especially genericsare only helpful for types, not for parametersQuantity does not help. Personally, I would approach this problem in a different way:
func decorator(fp func()) { // do stuff fp() // do other stuff }
Call it using a closure:
var res sometype decorator(func() { res = originalfunc(arg1, arg2, arg3) })
It's a bit verbose at the call site, but it works absolutely for any function signature - any number or type of parameters and any number or type of return value. It even works with method calls. It sounds like you might be doing some logic based on the results, so you could also:
func Decorator(fp func() bool) { // do stuff res := fp() // do other stuff } var res someType Decorator(func() bool { res = originalFunc(arg1, arg2, arg3) return res })
It will still work even if the decorated function has more return values than that bool, or if you want to use other logic (for example, if it returns an error you want to convert to a bool using !=). nil
).
The above is the detailed content of Is this a valid use case for generics?. For more information, please follow other related articles on the PHP Chinese website!