Custom ID

Ladda needs an unique identifier for every EntityValue to cache it. By default Ladda assumes that every EntityValue has this identifier as a property id. Of course, we realize that this is not always true, so you can configure what to use as a identifier in a couple of ways.

Default ID Property

Sometimes you have identifiers, but they are called _id, objectID or something else. In this case you will need to tell Ladda about this, for example, if your default ID property is objectID, you could configure this as:

const config = {
    user: {
        ttl: 60,
        api: userApi
    },
    __config: {
        idField: 'objectID'
    }
};

This will tell Ladda to always pick objectID rather than id as the ID for the EntityValue.

Override ID Property for ApiFunction

There are always exceptions. Maybe most of your API use objectID as the ID property. But a few parts of your API use _id. This can easily be handled by configuring it on your lower level ApiFunctions:

getUser.operation = 'READ';
getUser.idFrom = user => user._id;
function getUser(id) {
    return performGetRequest('/api/user', id);
}

This will tell Ladda to always pick _id rather than objectID as the ID for the EntityValue.

No ID at All

Finally, you might not have any ID. For example, if your endpoint uses a custom format. In general, we recommend you to put some effort into getting an ID. For instance, by updating your backend, preprocessing your response in your ApiFunction or to use idFrom with a function on your ApiFunction.

However, if you do not plan to manipulate the response (eg. update and post the updated version to your backend), then you can treat it as a BlobValue. This means that Ladda will not try to do anything clever except of caching the value and returning it if you call the same method with the same arguments on a READ operation. Basically, it will be equivalent to the good ol' if (inCache) { return cachedValue; } else { val = loadValue; putInCache(val); return val;}. This removes the requirement of the response having an ID, the ID will be created by Ladda by serializing the arguments of the called ApiFunction. To enable this, configure your ApiFunction as:

getHackernewsArticles.operation = 'READ';
getHackernewsArticles.idFrom = 'ARGS';
function getHackernewsArticles(query) {
    return performGetRequst(HACKERNEWS_SEARCH_URL, query);
}

The ARGS will work regardless of what Hackernews returns. And calling the ApiFunction with the same query twice will give you a cached result. However, if you had a updateHackernewsArticle ApiFunction, it would not update the cached result later returned by getHackernewsArticles. You would need to invalidate getHackernewsArticles to prevent stale data:

updateHackernewsArticle.operation = 'UPDATE';
updateHackernewsArticle.invalidates = ['getHackernewsArticles'];
function updateHackernewsArticle(newArticle) {
    return performPutRequest(HACKERNEWS_ARTICLE_URL, newArticle);
}

getHackernewsArticles.operation = 'READ';
getHackernewsArticles.idFrom = 'ARGS';
function getHackernewsArticles(query) {
    return performGetRequst(HACKERNEWS_SEARCH_URL, query);
}