Let's say I have a test for a JSON API that looks like this:

#[test]
fn create_session_success() {
    let conn = DB_POOL.get().unwrap();
    let mut user = NewUser::new("foo@bar.com", "correcthorsebatterystaple", "Foo Bar").save(&*conn);
    user.status = User::STATUS_ACTIVE;
    user.update(&*conn);

    let rocket = api::build_rocket();
    let req_body = json!({
        "email": "foo@bar.com",
        "password": "correcthorsebatterystaple"
    });
    let mut req = MockRequest::new(Method::Post, "/api/sessions")
        .header(ContentType::JSON)
        .body(&req_body.to_string());
    let mut response = req.dispatch_with(&rocket);
    assert_eq!(response.status(), Status::Ok);
    let resp_body: HashMap<String, Value> =
        serde_json::from_str(&response.body().and_then(|b| b.into_string()).unwrap()).unwrap();
    assert!(resp_body.contains_key(&"token".to_string()));
}

What would be the benefit to using strings other than "foo" and "bar", if I just want to ensure that all of the fields are filled in and valid? I can see the benefit for variable names--the code becomes harder to understand if I rename user to foo--but I don't see the benefit for strings.

Lint for denylisted-strings, similar to denylisted-names #1624

Description

clarfontheyopenedon Mar 15, 2017

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

clarfonthey
openedon Mar 15, 2017