How can I avoid a function name clash while implementing a trait?

Question

I have a struct with an implementation that has a function which accesses the private state of the struct.

struct Example {...}

impl Example {
    fn test(&self) -> .. {...}
}

Somewhere else in another module there exists another trait:

trait ExampleTrait {
    fn test(&self) -> .. {...}
}

Now I would like to implement ExampleTrait for the Example struct and to forward the test method to the test impl for the struct.

The following code:

impl ExampleTrait for Example {
    fn test(&self) -> .. {
        self.test()
    }
}

Is obviously an infinite recursive call. I cannot just repeat the body of the original test as I do not have access to the private state of Example here.

Is there another way to do this except for renaming one function or making fields in Example public?

score 11 · Accepted Answer · edited Mar 16 '18 at 14:24

11

You can use the fully-qualified syntax to disambiguate which method is to be used:

impl ExampleTrait for Example {
    fn test(&self) {
        Example::test(self) // no more ambiguity
    }
}

edited Mar 16 '18 at 14:24

Shepmaster

388,571
95
1,107
1,366

answered Mar 16 '18 at 11:51

ljedrz

20,316
4
69
97

is it possible to avoid adding this method stub completely? for instance if a trait has 10 functions which are already derived from another macro/trait, how can I avoid adding this boiler plate? (in other words just add a `impl SomeTrait for MyType {// leave empty since function names coincide and are already implemented}` – Avba Jan 25 '22 at 11:39

How can I avoid a function name clash while implementing a trait?

1 Answers1