Is it possible to switch to a new View without using NavigationStack or NavigationLink or NavigationView?
I know I can do it with a Bool, which either shows the second view, or the first, and then toggles the Bool.
But can't I do something like this? Which obviously doesn't work.
struct BasicButton: View {
var buttonLabel = "Create User"
var body: some View {
Button {
CreateUser() //another SwiftUI view, not a function
} label: {
Text(buttonLabel)
}
}
}
But can't I do something like this?
No, you can't, for reasons that are a bit more philosophical than technical.
In "normal" Swift code (non-SwifUI, non-result-builder), there is an execution flow sequence over time. One statement executes after another.
In SwifUI/result-builder "code", time isn't a factor. Instead, the code describes the arrangement and relationships between things in a time-neutral sense.
When you use a button to trigger an action or change of state, you're necessarily talking about time sequence: press the button, then do something. That's why the code inside Button's closure is regular Swift code, not result-builder code.
The only way to bridge the two worlds is the solution you already know: use (e.g.) a Boolean @state property to choose between 2 views.
Note that even declarative SwiftUI code can have time-related implications. Your Button usage outside of the action closure is declarative, but clearly the button action will happen "later", so time is a factor. You just don't control the "when" directly.
I accept your answer, but that shouldn't be the answer. As dangerous as certain things are, i.e. calling a null pointer, we should be allowed to design as we like.
But can't I do something like this?
No, you can't, for reasons that are a bit more philosophical than technical.
In "normal" Swift code (non-SwifUI, non-result-builder), there is an execution flow sequence over time. One statement executes after another.
In SwifUI/result-builder "code", time isn't a factor. Instead, the code describes the arrangement and relationships between things in a time-neutral sense.
When you use a button to trigger an action or change of state, you're necessarily talking about time sequence: press the button, then do something. That's why the code inside Button's closure is regular Swift code, not result-builder code.
The only way to bridge the two worlds is the solution you already know: use (e.g.) a Boolean @state property to choose between 2 views.
Note that even declarative SwiftUI code can have time-related implications. Your Button usage outside of the action closure is declarative, but clearly the button action will happen "later", so time is a factor. You just don't control the "when" directly.
I accept your answer, but that shouldn't be the answer. As dangerous as certain things are, i.e. calling a null pointer, we should be allowed to design as we like.
You just don't control the "when" directly.
Yes, but we can have some control with a dispatch:
Button {
// CreateUser() //another SwiftUI view, not a function
DispatchQueue.main.asyncAfter(deadline: .now() + 3) {
showCreate.toggle()
}
} label: {
Text(buttonLabel)
}
I'd call that "indirectly". 😄
The distinction I was trying to make is that the use of Button is itself in a view-builder (aka SwiftUI declarative code), but the inside of the Button closure is procedural code, not declarative code, so yes it absolutely has a time sequence and the ability to control timing.
Keep in mind that the OP asked about handling the timing in SwiftUI code, without toggling the state variable "when…".
