Golang嵌入式结构

Question

I am a little confused with a behavior of Go regarding variable 'overriding' in embedded structure.

First situation If a child structure embeds a parent structure containing a field Attr, I can access the value of Attr indifferently with child.Attr or child.parent.Attr. Here is an example :

package main

import (
    "fmt"
    "encoding/json"
)

type parent struct {
    Attr    int `json:"attr"`
}

type child struct {
    parent
}

func main() {
    var c child
    json.Unmarshal([]byte(`{"i": 1}`), &c)
    fmt.Println(c.Attr)
    fmt.Println(c.parent.Attr)
}

Second situation However, if the child structure contains itself a field named Attr, those two fields are different and can be accessed separately, as shows the following example :

package main

import (
    "fmt"
    "encoding/json"
)

type parent struct {
    Attr    int `json:"attr"`
}

type child struct {
    parent
    Attr    int
}

func main() {
    var c child
    json.Unmarshal([]byte(`{"attr": 1}`), &c)
    fmt.Println(c.Attr)
    fmt.Println(c.parent.Attr)
}

I am very surprised that this implicit behavior is allowed in golang. I would have expected the language to be stricter as it is on so many points. Besides, I wasn't able to find a clear specification about this. Is it just a side-effect or can I use that feature?

Answer 1

Golang specification actually states how embedded fields are resolved

A selector f may denote a field or method f of a type T, or it may refer to a field or method f of a nested embedded field of T. The number of embedded fields traversed to reach f is called its depth in T. The depth of a field or method f declared in T is zero. The depth of a field or method f declared in an embedded field A in T is the depth of f in A plus one.

Then...

For a value x of type T or *T where T is not a pointer or interface type, x.f denotes the field or method at the shallowest depth in T where there is such an f. If there is not exactly one f with shallowest depth, the selector expression is illegal.