I was wondering how you guys would solve the following problem

I have an actix api the following thoughts about my response model. it's a document based response (not field based resposne filtering atm although I am thinking about it)

• code (for easy switches/mappings in client)
• error message (so the custom client errors are optional)
• success branch (for generic data)
• error branch (for error data)

the first 2 parameters are just for the user to make decisions before accessing the data.

the reason I got 2 branches instead just a code + data setting is in theory you could get 100 okay things and 1 failed things. (debatable I know - not sure myself)

# models.rs


#[derive(Serialize, Deserialize, PartialEq)]
pub struct AppResponse<T=String, E=String> {
    pub status: i16,
    pub message: Option<String>,
    pub success: Option<T>,
    pub error: Option<E>,
}

impl <T>AppResponse<T> {

    pub fn success(success: T, message: String) -> AppResponse<T> {
        AppResponse {
            status: 1,
            message: Some(message.clone()),
            success: Some(success),
            error: None
        }
    }
}

impl <E>AppResponse<E> {

    pub fn fail(error: E, message: String) -> AppResponse<Option<String>, E> {
        AppResponse {
            status: 1,
            message: Some(message.clone()),
            success: None,
            error: Some(error),
        }
    }
}

impl <T, E>AppResponse<T, E> {

    pub fn mixed(success: T, error: E, message: String) -> AppResponse<T, E> {
        AppResponse {
            status: 1,
            message: Some(message.clone()),
            success: Some(success),
            error: Some(error),
        }
    }
}

This is how I currently solved it. I am pretty sure there is a better solution than just defaulting to string, probably defaulting to bool ... not sure. it works that's step one.

I am curious about different normalized response models. I should probably add partial equals and filter over map so a more matrix like approach for data. So go for it :) give me feedback :).