Client side TTL caching strategy for redux applications
This library provides an easy to use client side TTL caching strategy for redux applications.
Often, it is unnecessary to refetch data from an API if we know that it is unlikely to have changed within a certain period. By not fetching the data we are being friendly to mobile data users, limiting unnecessary API calls, and minimizing re-renders of applications due to state updates.
npm install -s redux-cache
Note: For a complete example, have a look at the example directory. You can clone this repo and actually start the example to see it working.
Add the store enhancer to your redux store. Example:
import { compose, applyMiddleware, createStore } from 'redux'
import { cacheEnhancer } from 'redux-cache'
const store = createStore(
reducer,
undefined,
compose(
applyMiddleware(...),
cacheEnhancer()
)
)
This will enhance your store so that whenever you dispatch the invalidateCache action, it automatically will invalidate the cache for any provided reducers. This means you do not have to bother with the boilerplate for every single reducer. Great for larger apps!
Included are a couple of utilities for setting up your reducer to make it "cache enabled".
Firstly, the DEFAULT_KEY
is what redux-cache
will use for storing and clearing cache values unless told to use an additional cache key. Secondly is the generateCacheTTL
function which will create a timestamp for you.
In your reducer:
// postsReducer.js
import { DEFAULT_KEY, generateCacheTTL } from "redux-cache";
const initialState = {
[DEFAULT_KEY]: null
// ...other keys
}
const postReducer = (state = initialState, action) => {
switch (action.type) {
case FETCH_POSTS_SUCCESS: {
return {
...state,
[DEFAULT_KEY]: generateCacheTTL(),
results: action.payload
};
}
default: {
return state;
}
}
}
We set up the key in the initial state and then on the successful fetch action we generate a new TTL time.
Before you fire off another get request, you should check if the cache is valid. A util fn - checkCacheValid()
- has been provided:
// postsActions.js
import { checkCacheValid } from "redux-cache";
export const getPosts = () => (dispatch, getState) => {
const isCacheValid = checkCacheValid(getState, "posts");
if (isCacheValid) { return null; }
dispatch({
type: FETCH_POSTS_REQUEST
});
axios.get("https://jsonplaceholder.typicode.com/posts")
.then((response) => {
dispatch({
type: FETCH_POSTS_SUCCESS,
payload: response.data,
});
})
.catch((error) => {
console.log('error: ', error);
dispatch({
type: FETCH_POSTS_FAILURE,
payload: error,
});
});
};
If the cache is valid, drop the action on the floor. This will stop any calls to the API being made.
If you need to invalidate the cache for a reducer there is a utility action provided - invalidateCache
. Just dispatch this with the reducer(s) you need to invalidate.
import { invalidateCache } from "redux-cache";
invalidateCache("posts");
cacheEnhancer([config])
DEFAULT_KEY
. The cacheKey to be updating in your reducers when invalidating the cache.generateCacheTTL([duration])
DEFAULT_DURATION_MS
. A number representing miliseconds. To be added to the current time.checkCacheValid(getState, reducerKey, [config])
redux-thunk
which will be used to get the application stateDEFAULT_KEY
. The cacheKey to be checking.invalidateCache(reducers)
DEFAULT_KEY
- "cacheUntil"DEFAULT_DURATION_MS
- 600000 (in milliseconds). 10 minutes.To run the tests:
npm run test