当前位置：文江博客知识库 en-US 文档 Learn JavaScript Asynchronous Choosing_the_right_approach

Choosing the right approach - Learn web development 编辑

To finish this module off, we'll provide a brief discussion of the different coding techniques and features we've discussed throughout, looking at which one you should use when, with recommendations and reminders of common pitfalls where appropriate. We'll probably add to this resource as time goes on.

Prerequisites:	Basic computer literacy, a reasonable understanding of JavaScript fundamentals.
Objective:	To be able to make a sound choice of when to use different asynchronous programming techniques.

Asynchronous callbacks

Generally found in old-style APIs, involves a function being passed into another function as a parameter, which is then invoked when an asynchronous operation has been completed, so that the callback can in turn do something with the result. This is the precursor to promises; it's not as efficient or flexible. Use only when necessary.

Useful for...
Single delayed operation	Repeating operation	Multiple sequential operations	Multiple simultaneous operations
No	Yes (recursive callbacks)	Yes (nested callbacks)	No

Code example

An example that loads a resource via the XMLHttpRequest API (run it live, and see the source):

function loadAsset(url, type, callback) {
  let xhr = new XMLHttpRequest();
  xhr.open('GET', url);
  xhr.responseType = type;

  xhr.onload = function() {
    callback(xhr.response);
  };

  xhr.send();
}

function displayImage(blob) {
  let objectURL = URL.createObjectURL(blob);

  let image = document.createElement('img');
  image.src = objectURL;
  document.body.appendChild(image);
}

loadAsset('coffee.jpg', 'blob', displayImage);

Pitfalls

Nested callbacks can be cumbersome and hard to read (i.e. "callback hell").
Failure callbacks need to be called once for each level of nesting, whereas with promises you can just use a single .catch() block to handle the errors for the entire chain.
Async callbacks just aren't very graceful.
Promise callbacks are always called in the strict order they are placed in the event queue; async callbacks aren't.
Async callbacks lose full control of how the function will be executed when passed to a third-party library.

Browser compatibility

Really good general support, although the exact support for callbacks in APIs depends on the particular API. Refer to the reference documentation for the API you're using for more specific support info.

Further information

Introducing asynchronous JavaScript, in particular Async callbacks

setTimeout()

setTimeout() is a method that allows you to run a function after an arbitrary amount of time has passed.

Useful for...
Single delayed operation	Repeating operation	Multiple sequential operations	Multiple simultaneous operations
Yes	Yes (recursive timeouts)	Yes (nested timeouts)	No

Code example

Here the browser will wait two seconds before executing the anonymous function, then will display the alert message (see it running live, and see the source code):

let myGreeting = setTimeout(function() {
  alert('Hello, Mr. Universe!');
}, 2000)

Pitfalls

You can use recursive setTimeout() calls to run a function repeatedly in a similar fashion to setInterval(), using code like this:

let i = 1;
setTimeout(function run() {
  console.log(i);
  i++;

  setTimeout(run, 100);
}, 100);

There is a difference between recursive setTimeout() and setInterval():

Recursive setTimeout() guarantees at least the specified amount of time (100ms in this example) will elapse between the executions; the code will run and then wait 100 milliseconds before it runs again. The interval will be the same regardless of how long the code takes to run.
With setInterval(), the interval we choose includes the time taken to execute the code we want to run in. Let's say that the code takes 40 milliseconds to run — the interval then ends up being only 60 milliseconds.

When your code has the potential to take longer to run than the time interval you’ve assigned, it’s better to use recursive setTimeout() — this will keep the time interval constant between executions regardless of how long the code takes to execute, and you won't get errors.

Browser compatibility

BCD tables only load in the browser

Further information

Cooperative asynchronous JavaScript: Timeouts and intervals, in particular setTimeout()
setTimeout() reference

setInterval()

setInterval() is a method that allows you to run a function repeatedly with a set interval of time between each execution. Not as efficient as requestAnimationFrame(), but allows you to choose a running rate/frame rate.

Useful for...
Single delayed operation	Repeating operation	Multiple sequential operations	Multiple simultaneous operations
No	Yes	No (unless they are the same)	No

Code example

The following function creates a new Date() object, extracts a time string out of it using toLocaleTimeString(), and then displays it in the UI. We then run it once per second using setInterval(), creating the effect of a digital clock that updates once per second (see this live, and also see the source):

function displayTime() {
   let date = new Date();
   let time = date.toLocaleTimeString();
   document.getElementById('demo').textContent = time;
}

const createClock = setInterval(displayTime, 1000);

Pitfalls

The frame rate isn't optimized for the system the animation is running on, and can be somewhat inefficient. Unless you need to choose a specific (slower) framerate, it is generally better to use requestAnimationFrame().

Browser compatibility

BCD tables only load in the browser

Further information

Cooperative asynchronous JavaScript: Timeouts and intervals, in particular setInterval()
setInterval() reference

requestAnimationFrame()

requestAnimationFrame() is a method that allows you to run a function repeatedly, and efficiently, at the best framerate available given the current browser/system. You should, if at all possible, use this instead of setInterval()/recursive setTimeout(), unless you need a specific framerate.

Useful for...
Single delayed operation	Repeating operation	Multiple sequential operations	Multiple simultaneous operations
No	Yes	No (unless it is the same one)	No

Code example

A simple animated spinner; you can find this example live on GitHub (see the source code also):

const spinner = document.querySelector('div');
let rotateCount = 0;
let startTime = null;
let rAF;

function draw(timestamp) {
    if(!startTime) {
        startTime = timestamp;
    }

    rotateCount = (timestamp - startTime) / 3;

    if(rotateCount > 359) {
        rotateCount %= 360;
    }

    spinner.style.transform = 'rotate(' + rotateCount + 'deg)';

    rAF = requestAnimationFrame(draw);
}

draw();

Pitfalls

You can't choose a specific framerate with requestAnimationFrame(). If you need to run your animation at a slower framerate, you'll need to use setInterval() or recursive setTimeout().

Browser compatibility

BCD tables only load in the browser

Further information

Promises

Promises are a JavaScript feature that allows you to run asynchronous operations and wait until it is definitely complete before running another operation based on its result. Promises are the backbone of modern asynchronous JavaScript.

Useful for...
Single delayed operation	Repeating operation	Multiple sequential operations	Multiple simultaneous operations
No	No	Yes	See `Promise.all()`, below

Code example

The following code fetches an image from the server and displays it inside an <img> element; see it live also, and see also the source code:

fetch('coffee.jpg')
.then(response => response.blob())
.then(myBlob => {
  let objectURL = URL.createObjectURL(myBlob);
  let image = document.createElement('img');
  image.src = objectURL;
  document.body.appendChild(image);
})
.catch(e => {
  console.log('There has been a problem with your fetch operation: ' + e.message);
});

Pitfalls

Promise chains can be complex and hard to parse. If you nest a number of promises, you can end up with similar troubles to callback hell. For example:

remotedb.allDocs({
  include_docs: true,
  attachments: true
}).then(function (result) {
  let docs = result.rows;
  docs.forEach(function(element) {
    localdb.put(element.doc).then(function(response) {
      alert("Pulled doc with id " + element.doc._id + " and added to local db.");
    }).catch(function (err) {
      if (err.name == 'conflict') {
        localdb.get(element.doc._id).then(function (resp) {
          localdb.remove(resp._id, resp._rev).then(function (resp) {
// et cetera...

It is better to use the chaining power of promises to go with a flatter, easier to parse structure:

remotedb.allDocs(...).then(function (resultOfAllDocs) {
  return localdb.put(...);
}).then(function (resultOfPut) {
  return localdb.get(...);
}).then(function (resultOfGet) {
  return localdb.put(...);
}).catch(function (err) {
  console.log(err);
});

or even:

remotedb.allDocs(...)
.then(resultOfAllDocs => {
  return localdb.put(...);
})
.then(resultOfPut => {
  return localdb.get(...);
})
.then(resultOfGet => {
  return localdb.put(...);
})
.catch(err => console.log(err));

That covers a lot of the basics. For a much more complete treatment, see the excellent We have a problem with promises, by Nolan Lawson.

Browser compatibility

BCD tables only load in the browser

Further information

Promise.all()

A JavaScript feature that allows you to wait for multiple promises to complete before then running a further operation based on the results of all the other promises.

Useful for...
Single delayed operation	Repeating operation	Multiple sequential operations	Multiple simultaneous operations
No	No	No	Yes

Code example

The following example fetches several resources from the server, and uses Promise.all() to wait for all of them to be available before then displaying all of them — see it live, and see the source code:

function fetchAndDecode(url, type) {
  // Returning the top level promise, so the result of the entire chain is returned out of the function
  return fetch(url).then(response => {
    // Depending on what type of file is being fetched, use the relevant function to decode its contents
    if(type === 'blob') {
      return response.blob();
    } else if(type === 'text') {
      return response.text();
    }
  })
  .catch(e => {
    console.log(`There has been a problem with your fetch operation for resource "${url}": ` + e.message);
  });
}

// Call the fetchAndDecode() method to fetch the images and the text, and store their promises in variables
let coffee = fetchAndDecode('coffee.jpg', 'blob');
let tea = fetchAndDecode('tea.jpg', 'blob');
let description = fetchAndDecode('description.txt', 'text');

// Use Promise.all() to run code only when all three function calls have resolved
Promise.all([coffee, tea, description]).then(values => {
  console.log(values);
  // Store each value returned from the promises in separate variables; create object URLs from the blobs
  let objectURL1 = URL.createObjectURL(values[0]);
  let objectURL2 = URL.createObjectURL(values[1]);
  let descText = values[2];

  // Display the images in <img> elements
  let image1 = document.createElement('img');
  let image2 = document.createElement('img');
  image1.src = objectURL1;
  image2.src = objectURL2;
  document.body.appendChild(image1);
  document.body.appendChild(image2);

  // Display the text in a paragraph
  let para = document.createElement('p');
  para.textContent = descText;
  document.body.appendChild(para);
});

Pitfalls

If a Promise.all() rejects, then one or more of the promises you are feeding into it inside its array parameter must be rejecting, or might not be returning promises at all. You need to check each one to see what they returned.

Useful for...
Single delayed operation	Repeating operation	Multiple sequential operations	Multiple simultaneous operations
No	No	Yes	Yes (in combination with `Promise.all()`)

Code example

The following example is a refactor of the simple promise example we saw earlier that fetches and displays an image, written using async/await (see it live, and see the source code):

async function myFetch() {
  let response = await fetch('coffee.jpg');
  let myBlob = await response.blob();

  let objectURL = URL.createObjectURL(myBlob);
  let image = document.createElement('img');
  image.src = objectURL;
  document.body.appendChild(image);
}

myFetch();

Pitfalls

You can't use the await operator inside a non-async function, or in the top level context of your code. This can sometimes result in an extra function wrapper needing to be created, which can be slightly frustrating in some circumstances. But it is worth it most of the time.
Browser support for async/await is not as good as that for promises. If you want to use async/await but are concerned about older browser support, you could consider using the BabelJS library — this allows you to write your applications using the latest JavaScript and let Babel figure out what changes if any are needed for your user’s browsers.