Not checking for nil might lead to nil pointer dereference

[rzajac]

https://github.com/gorilla/schema/blob/8285576f31afd6804df356a38883f4fa05014373/encoder.go#L167-L173

The f should be checked for nil before it is used in closure.

[zak905]

Hi @rzajac, do you have an example that triggers a panic ? I think that if f is nil than v will be nil as well, and therefore the closure will return always "null" value, so f(v.Elem()) statement will not be reached

I tried to trigger the case you mentionned but I was not able to

func TestNilPointerDeference(t *testing.T) {
    type C struct {
        R int `schema:"R"`
    }

    type B struct {
        C *C `schema:"C"`
    }

    b := B{C: (*C)(nil)}

    encoder := NewEncoder()

    vals := url.Values{}

    if err := encoder.Encode(b, vals); err != nil {
        t.Fatalf(err.Error())
    }
}

The test is passing without errors.

[zak905]

@elithrar this can be closed I think

[jaitaiwan]

Thanks heaps @zak905 closing

[zak905]

sure thing!