I've seen the argument for separating actions and reducers because they have a many-to-many relationship.
I don't think that actually applies in Redux though. Because there's only 1 datastore, actions to reducers should be 1-to-many.
Typically reducers apply to a specific change for a specific datastore.
MY_ACTION = "MY_ACTION"
function reducer(state, action) {
switch(action.type) {
case MY_ACTION: // stuff with my action to create new state
default: return state
}
}
We can combine multiple reducers with combineReducers
so why not define the handler for an action with the action itself.
For instance
class Action {
constructor(type) {
this.type = type
this.handlers = []
}
add_handler(handler) {
this.handlers += handler
}
get_reducer() {
reducer = combineReducers(this.handlers)
return (state, action) => {
if(action.type == this.type) {
return reducer(state, action)
}
return state
}
}
}
With the "ducks" pattern, we end up putting the main reducers in the same module as the action declaration.
Is there any reason to keep reducers + actions separate with redux?
The main reason for separating the action creators from the reducer function is that the reducer function must be a pure function. If you wanted to do something in an action creator, like an asynchronous API call for instance, then you could not put this in the reducer. There's a great explanation on this here.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With