Reputation: 24679
My question relates to redux and more specifically how to handle errors/failures from within reducer functions. I am in reference to the ngrx example app (https://github.com/ngrx/example-app) and the way it handle errors/failures.
Here is the reducer function I am referring to:
export function reducer(state = initialState, action: collection.Actions): State {
switch (action.type) {
case collection.ActionTypes.LOAD: {
return Object.assign({}, state, {
loading: true
});
}
case collection.ActionTypes.LOAD_SUCCESS: {
const books = action.payload;
return {
loaded: true,
loading: false,
ids: books.map(book => book.id)
};
}
case collection.ActionTypes.ADD_BOOK_SUCCESS:
case collection.ActionTypes.REMOVE_BOOK_FAIL: {
const book = action.payload;
if (state.ids.indexOf(book.id) > -1) {
return state;
}
return Object.assign({}, state, {
ids: [ ...state.ids, book.id ]
});
}
case collection.ActionTypes.REMOVE_BOOK_SUCCESS:
case collection.ActionTypes.ADD_BOOK_FAIL: {
const book = action.payload;
return Object.assign({}, state, {
ids: state.ids.filter(id => id !== book.id)
});
}
default: {
return state;
}
}
}
Can someone please explain the necessity for dealing with those two actions from within the reducer function:
REMOVE_BOOK_FAIL
ADD_BOOK_FAIL
For instance why remove the book from the state (in the case of the ADD_BOOK_FAIL
action)?
If the add book action has failed, then the book is not present in the store. Is it?
Upvotes: 0
Views: 331
Reputation: 3844
Maybe it's the naming used that makes it a red herring, my guess is that ADD_BOOK_FAIL could be in use somewhere else for a different use case as a fall back mechanism.
I agree the way you describe it this doesnt make sense the developer did it for this reason.
Upvotes: 1