I'm working on a Node app with Express. I'm chaining several http calls to data api's, each dependent on the previous req's responses.
It's all working except the last call. The last call needs to happen multiple times before the page should render.
Searching has turned up excellent examples of how to chain, but not make a call to the same API (or HTTP GET, data endpoint, etc.) with different params each time.
I'm trying to do something like this: Using a generator to call an API multiple times and only resolve when all requests are finished?
var getJSON = (options, fn) => {
.....
}
router.route("/")
.get((req, res) => {
var idArray = [];
var results = [];
getJSON({
.... send params here, (result) => {
//add response to results array
results.push(result);
//create var for data nodes containing needed id params for next call
let group = result.groupsList;
//get id key from each group, save to idArray
for(i=0;i<groups.length;i++){
idArray.push(groups[I].groupId);
}
//use id keys for params of next api call
dataCallback(idArray);
});
function dataCallback(myArray){
// number of ID's in myArray determine how many times this API call must be made
myArray.forEach(element => {
getJSON({
.... send params here, (result) => {
results.push(result);
});
// put render in callback so it will render when resolved
}, myRender());
};
function myRender() {
res.render("index", { data: results, section: 'home'});
}
})
I learned the problem with the above code.
You can call functions that are outside of the express route, but you can't have them inside the route.
You can't chain multiple data-dependent calls, not in the route.
Anything inside route.get or route.post should be about the data, paths, renders, etc.
This means either using an async library (which I found useless when trying to build a page from multiple data sources, with data dependent on the previous response), or having an additional js file that you call (from your web page) to get, handle and model your data like here: Using a generator to call an API multiple times and only resolve when all requests are finished You could also potentially put it in your app or index file, before the routes.
(It wasn't obvious to me where that code would go, at first. I tried putting it inside my router.post. Even though the documentation says "Methods", it didn't click for me that routes were methods. I hadn't really done more than very basic routes before, and never looked under the hood.)
I ended up going with a third option. I broke up the various API calls in my screen so that they are only called when the user clicks on something that will need more data, like an accordion or tab switch.
I used an XMLHttpRequest() from my web page to call my own front-end Node server, which then calls the third party API, then the front-end Node server responds with a render of my pug file using the data the API provided. I get html back for my screen to append.
In page:
callFEroutetoapi(_postdata, _route, function (_newdata){
putData(_newdata);
});
function putData(tData){
var _html = tData;
var _target = document.getElementById('c-playersTab');
applyHTML(_target, _html);
}
function callFEroutetoapi(data, path, fn){
//url is express route
var url = path;
var xhr = new XMLHttpRequest();
console.log('data coming into xhr request: ', data);
//xhr methods must be in this strange order or they don't run
xhr.onload = function(oEvent) {
if(xhr.readyState === xhr.DONE) {
//if success then send to callback function
if(xhr.status === 200) {
fn(xhr.response);
// ]console.log('server responded: ', xhr.response);
}
else {
console.log("Something Died");
console.log('xhr status: ', xhr.status);
}
}
}
xhr.onerror = function (){console.log('There was an error.', xhr.status);}
xhr.open("POST", url, true);
xhr.setRequestHeader("Content-Type", "application/json;charset=UTF-8");
xhr.send(JSON.stringify(data));
}
It adds an extra layer, but was necessary to show the latest, frequently changing data. It's also reusable which is better for a multiscreen web app. If there were fewer views (completely different screens and co-dependent datasets), a more centralized model.js file mentioned above would work better.
Related
I'm building an app with Next.js... we have 100k+ pages and content changes daily, so using SSR and getServerSideProps.
Some of our data is coming from a headless CMS provider that charges by the request. I'd like to cache the API responses from this server for 24hrs.
What is the best way of going about this?
Is there a common library most folks use to do this?
Just looking for suggestions of approaches I should investigate (or great examples of how to do this).
I used this npm package:
https://www.npmjs.com/package/memory-cache
And then something like this:
import cacheData from "memory-cache";
async function fetchWithCache(url, options) {
const value = cacheData.get(url);
if (value) {
return value;
} else {
const hours = 24;
const res = await fetch(url, options);
const data = await res.json();
cacheData.put(url, data, hours * 1000 * 60 * 60);
return data;
}
}
Then if you want to fetch something with using the cache just call this function. Or it can be used as a midware in the requests. It checks if the data is already in the cache and returns it, or if not - it puts the data into the cache under the key. The key can be anything, I am using the url for instance.
In addition to Tobias Lins' answer:
At least if deploying on Vercel, you can use set Cache-Control headers in getStaticProps, getServerSideProps, API routes, etc to cache responses on Vercel's edge network. This solution does not require any additional dependencies and very minimal code.
api route example - source Vercel
// pages/api/user.js
export default function handler(req, res) {
res.setHeader('Cache-Control', 's-maxage=86400');
res.status(200).json({ name: 'John Doe' });
}
Example in getServerSideProps - Source NextJS
// This value is considered fresh for ten seconds (s-maxage=10).
// If a request is repeated within the next 10 seconds, the previously
// cached value will still be fresh. If the request is repeated before 59 seconds,
// the cached value will be stale but still render (stale-while-revalidate=59).
//
// In the background, a revalidation request will be made to populate the cache
// with a fresh value. If you refresh the page, you will see the new value.
export async function getServerSideProps({ req, res }) {
res.setHeader(
'Cache-Control',
'public, s-maxage=10, stale-while-revalidate=59'
)
return {
props: {},
}
}
I believe you'd want to use:
res.setHeader('Cache-Control', 's-maxage=1440000')
Here are some other useful links for caching on Vercel:
https://vercel.com/docs/concepts/functions/edge-caching
https://vercel.com/docs/concepts/edge-network/overview
https://vercel.com/docs/concepts/edge-network/caching
https://vercel.com/docs/concepts/edge-network/headers
For your specific case, you also may want to look into using getStaticPaths with getStaticProps. You can use fallback: true on getStaticPaths to only build pages when they're visited (you can still build your post popular pages at initial build time).
https://nextjs.org/docs/basic-features/data-fetching#the-fallback-key-required
I know this is an old post, but for others googling (at least those deploying on Vercel), these solutions should help where revalidate in getStaticProps does not.
You could use getStaticProps from Next.js for SSG
They currently have a revalidate property that you can return, that defines how often the content should be re-fetched.
Take a look here:
https://nextjs.org/blog/next-9-5#stable-incremental-static-regeneration
This is how we did it without any 3rd party libraries, as in our use-case we only had to cache a relatively smaller amount of global data(header/footer menus) which was shared across the site.
The data was coming from a CMS via GraphQL.
We ran an async method getGlobalData on each page from on getStaticProps method and then returned the cached data to the page component via props.
import fs from 'fs';
import path from 'path';
// Cache files are stored inside ./next folder
const CACHE_PATH = path.join(__dirname, 'globalData.json');
export default async function getGlobalData() {
let cachedData;
// #1 - Look for cached data first
try {
cachedData = JSON.parse(fs.readFileSync(CACHE_PATH, 'utf8'));
} catch (error) {
console.log('❌ CACHE NOT INITIALIZED');
}
// #2 - Create Cache file if it doesn't exist
if (!cachedData) {
// Call your APIs to-be-cached here
const data = await fetchGlobalData();
// Store data in cache files
// this always rewrites/overwrites the previous file
try {
await fs.writeFileSync(
CACHE_PATH,
JSON.stringify(data),
err =>throw err
);
console.log('💾 CACHE FILE WRITTEN SUCCESSFULLY');
} catch (error) {
console.log('❌ ERROR WRITING MEMBERS CACHE TO FILE\n', error);
}
cachedData = data;
}
return cachedData;
}
Call getGlobalData method from getStaticProps.
export async function getStaticProps({ preview = false }) {
const globalData = await getGlobalData();
// call other page-specific/non-shared APIs here
// ...
return { props: { globalData } };
}
References
https://flaviocopes.com/nextjs-cache-data-globally/
Note if you get an error saying fs or path is unknown or invalid, then please understand that, the above code is supposed to be running or referenced "serverside" i.e only inside getStaticProps or getServerSideProps. If you import and reference it "browser-side", say somewhere inside your components or on the page (other than methods mentioned above), then you will get an error, as there is no filesystem(fs) or path modules on browser. They are only available on node.
I have a React component which is making an API call on init on the client side. I don't want to hard-code my API key (god forbid in the repo), and it's not much better to put it in Meteor.settings.public since that can just be looked up in the console. I want to keep it in Meteor.settings, but then it's invisible to the client. I've tried using a method, but although it appears to work on the server the method call returns undefined on the client.
On the server:
Meteor.methods({
getFileStackAPIKey: function () {
if (Meteor.settings.fileStackAPIKey) {
console.log(Meteor.settings.fileStackAPIKey) // returns: [fileStackAPIKey] correctly
return Meteor.settings.fileStackAPIKey
}
else {
return {message: "Configure Meteor.settings.fileStackAPIKey to connect to FileStack."}
}
}});
On the client:
console.log(Meteor.call('getFileStackAPIKey')); // returns: undefined
I tried to use ReactiveVar and again it set it on the server but was inaccessible on the client. I have the feeling that I'm missing something obvious. Specifically, what I'm trying to make work is FileStack. Their example code shows the API key hard-coded inline. As does the official FileStack React package. This just doesn't seem like a good idea.
It has to do with callbacks. The method result will be in the callback, so what I needed to do on the client was more like this:
Meteor.call('getFileStackAPIKey', (err, res) => {
console.log("FileStack API Key: " + res);
});
But because what I really wanted to do was pass it into the FileStack init (again, on the client side), so I needed to put the following in the constructor for the FileStack object:
// "this" is the FileStack object we're constructing
const fileStackObj = this;
Meteor.call('getFileStackAPIKey', (err, apiKey) => {
// here we're inside the callback, so we have the resulting API key
const client = filestack.init(apiKey, clientOptions);
// these are synchronous actions dependent on the existence of "client"
// that we could not do outside of the callback
fileStackObj.state = {
client,
picker: action === 'pick' ? client.picker({ ...actionOptions, onUploadDone: fileStackObj.onFinished }) : null,
};
fileStackObj.onFinished = fileStackObj.onFinished.bind(fileStackObj);
fileStackObj.onFail = fileStackObj.onFail.bind(fileStackObj);
});
I am developing a react js application and we are using a promise based library axios for calling APIs.
Now, in the initial part of application, user gets a login page, when the login is successful, we contact different systems to retrieve some extra information about user.
axios
.get('url to authentication endpoint') // 1st call
.then(response => {
// if login is successful
// 1. retrieve the user preferences like, on the customised screens what fields user wanted to see
axios.get('user preference endpoint') // 2nd call
// 2. send a request to one more external systems, which calculates what user can see and not based on LDAP role
axios.get('role calculation endpoint') // 3rd call
})
.catch(error => {
})
Now I can see that I can use
axios.all()
for second and third call, but with promised based client, how to chain first and second call? To retrieve user preferences, I have to wait for user to be authenticated.
How to chain this calls in a promise based way, rather than callback style?
as mentioned in the thread for this Github issue, axios() and axios.all() return Promise objects which can be chained however you see fit:
axios.get('/auth')
.then(function(response) {
return axios.all([ axios.get('/preferences'), axios.get('/roles') ]);
})
.then(function(responses) {
const [
preferencesResponse,
rolesResponse
] = responses;
// do more things
})
.catch(function(error) {
console.log(error);
});
Dan O's answer is very good and it works perfectly but it's much readable using async/await although it's also working with promises under the hoood
async yourReactClassFunction(){
try{
let getAuth = await axios.get('/auth');
//if login not successful return;
let result = await Promise.all([axios.get('/preferences'), axios.get('/roles')]);
//Do whatever with the results.
}catch(e){
//TODO error handling
}
}
Although it's the same thing, 'feels' more readable in my very subjective opinion
In my react App I have a input element. The search query should be memoized, which means that if the user has previously searched for 'John' and the API has provided me valid results for that query, then next time when the user types 'Joh', there should be suggestion for the user with the previously memoized values(in this case 'John' would be suggested).
I am new to react and am trying caching for the first time.I read a few articles but couldn't implement the desired functionality.
You don't clarify which API you're using nor which stack; the solution would vary somewhat depending on if you are using XHR requests or something over GraphQL.
For an asynchronous XHR request to some backend API, I would do something like the example below.
Query the API for the search term
_queryUserXHR = (searchTxt) => {
jQuery.ajax({
type: "GET",
url: url,
data: searchTxt,
success: (data) => {
this.setState({previousQueries: this.state.previousQueries.concat([searchTxt])
}
});
}
You would run this function whenever you want to do the check against your API. If the API can find the search string you query, then insert that data into a local state array variable (previousQueries in my example).
You can either return the data to be inserted from the database if there are unknowns to your view (e.g database id). Above I just insert the searchTxt which is what we send in to the function based on what the user typed in the input-field. The choice is yours here.
Get suggestions for previously searched terms
I would start by adding an input field that runs a function on the onKeyPress event:
<input type="text" onKeyPress={this._getSuggestions} />
then the function would be something like:
_getSuggestions = (e) => {
let inputValue = e.target.value;
let {previousQueries} = this.state;
let results = [];
previousQueries.forEach((q) => {
if (q.toString().indexOf(inputValue)>-1) {
result.push(a);
}
}
this.setState({suggestions: results});
}
Then you can output this.state.suggestions somewhere and add behavior there. Perhaps some keyboard navigation or something. There are many different ways to implement how the results are displayed and how you would select one.
Note: I haven't tested the code above
I guess you have somewhere a function that queries the server, such as
const queryServer = function(queryString) {
/* access the server */
}
The trick would be to memorize this core function only, so that your UI thinks its actually accessing the server.
In javascript it is very easy to implement your own memorization decorator, but you could use existing ones. For example, lru-memoize looks popular on npm. You use it this way:
const memoize = require('lru-memoize')
const queryServer_memoized = memoize(100)(queryServer)
This code keeps in memory the last 100 request results. Next, in your code, you call queryServer_memoized instead of queryServer.
You can create a memoization function:
const memo = (callback) => {
// We will save the key-value pairs in the following variable. It will be our cache storage
const cache = new Map();
return (...args) => {
// The key will be used to identify the different arguments combination. Same arguments means same key
const key = JSON.stringify(args);
// If the cache storage has the key we are looking for, return the previously stored value
if (cache.has(key)) return cache.get(key);
// If the key is new, call the function (in this case fetch)
const value = callback(...args);
// And save the new key-value pair to the cache
cache.set(key, value);
return value;
};
};
const memoizedFetch = memo(fetch);
This memo function will act like a key-value cache. If the params (in our case the URL) of the function (fetch) are the same, the function will not be executed. Instead, the previous result will be returned.
So you can just use this memoized version memoizedFetch in your useEffect to make sure network request are not repeated for that particular petition.
For example you can do:
// Place this outside your react element
const memoizedFetchJson = memo((...args) => fetch(...args).then(res => res.json()));
useEffect(() => {
memoizedFetchJson(`https://pokeapi.co/api/v2/pokemon/${pokemon}/`)
.then(response => {
setPokemonData(response);
})
.catch(error => {
console.error(error);
});
}, [pokemon]);
Demo integrated in React
I have a service that does an http request to save some data. When the data comes from the backend I am doing some manipulation on the data and then return it so that controllers can use them. Something like:
public savePerson = (person: Model.IPerson): ng.IPromise<Model.IMiniPerson> => {
return this.api.persons.save({}, person).then((savedPerson) => {
this.enrichWithLookups(savedPerson);
var miniPerson = new Model.MiniPerson();
angular.extend(miniPerson, savedPerson);
miniPerson.afterLoad();
this.persons.unshift(miniPerson);
this.notifyOfChanges();
return miniPerson;
});
}
In order to clean up the code a bit and make it more testable I wanted to remove the private manipulation functions into decorating/intercepting services. Problem is I do not know how to hook on the promise data before the success function is executed and after it is returned.
For example enrichWithLookups must be applied first just after the data arrives and not after the miniPerson is returned.
you can create a local promise and call the "resolve" method when you have completed your operations on the http response. Look at the code down here:
public savePerson = (person: Model.IPerson): ng.IPromise<Model.IMiniPerson> => {
var waiter = $q.defer();
this.api.persons.save({}, person).then((savedPerson) => {
this.enrichWithLookups(savedPerson);
var miniPerson = new Model.MiniPerson();
angular.extend(miniPerson, savedPerson);
miniPerson.afterLoad();
this.persons.unshift(miniPerson);
this.notifyOfChanges();
waiter.resolve(miniPerson);
});
return waiter.promise;
}
I've wrote the code directly with angularjs, but I think that you can easily adapt it to fit your needs.
Bye.