CODEWITHSUNDEEP
gotestingnull
xD-saleem
xD-saleem

Reputation: 127

golang test assert the nil return from function is failing

I am trying to assert a function that returns a nil, but i'm trying to assert nil but the following assertion doesn't make sense. I am using the github.com/stretchr/testify/assert framework to assert

passes: assert.Equal(t, meta == nil, true)

fails: assert.Equal(t, meta, nil)

I am not sure why this makes any sense. could someone help please.

Method being tested:

type Metadata struct {
    UserId          string
}

func GetAndValidateMetadata(metadata map[string]string) (*Metadata, error) {
    userId _ := metadata["userId"]
    if userId == "" {
        return nil, errors.New("userId is undefined")
    }
    meta := Metadata{
        UserId:          userId,
    
    }
    return &meta, nil
}

testcase:

func TestValidation(t *testing.T) {
    metadata := map[string]string{
        "fakeUserId": "testUserId",
    }
    meta, err := GetAndValidateMetadata(metadata)

 assert.Equal(t, meta == nil, true) <--- passes
 assert.Equal(t, meta, nil) <--- fails
}

Upvotes: 6

Views: 10600

Answers (1)

Peter
Peter

Reputation: 31691

Interface values in Go store a type and a value of that type.

The equal method takes interface{} arguments, so in the second assertion you are comparing an interface value containing type information (*Metadata) and a nil value to an interface without type information and a nil value.

You can force the second interface to also contain a type like so:

assert.Equal(t, meta, (*Metadata)(nil))

See also "Why is my nil error value not equal to nil?" in the FAQ.

Upvotes: 12

Related Questions