150 Days of HTML

Chapter 10 | The integrity attribute

In this chapter we are moving onward and looking at the integrity attribute of the link element. To be clear, as with many of the other attributes we have discussed, the integrity attribute is not only used on the link element. For our immediate purpose though, we will focus on the link element.

When we talked about the crossorigin attribute I touched on the fact that these days very few sites load all of their dependencies from the same origin. We use a combination of third-party services, content delivery networks, etc., and pretty much trust that the resource we requested will be the one that is delivered.

With hacking techniques such as man-in-the-middle attacks, DNS spoofing, and DNS cache poisoning(among others), simply trusting is not enough, especially for critical applications that involve user data, payments, or other sensitive operations. This is where the integrity attribute can help. πŸ”

The integrity attribute is part of a larger standard known as subresource integrity. While TLS(Transport Layer Security)/SSL(Secure Soccer Layer) is concerned with validating that the server you connect to is the server you intended to connect to, it does not authenticate the content sent by the server. Subresource integrity is concerned with the latter of the two.

How-To πŸ‘©β€πŸ«

The details of this are beyond our current scope but, here is how you would use it in practice.

Let’s say we have the following CSS file: main.css

body {
  background-color: #212121;
  color: #fff;
  font: 18px/1.5 sans-serif;

We host this file on a CDN but wish to ensure that the content we get is the content we expect. Our first step is to generate a cryptographic hash from the above content. Here I will be using OpenSSL’s tools to generate the hash from the command line.

cat main.css | openssl dgst -sha384 -binary | openssl base64 -A | pbcopy

Let’s break the above down: πŸ•

  • We use the cat command to get the contents of the file
  • We pipe(send) the contents of the file to openSSL’s dgst function which generates a sha384 hash
  • We pipe the result of this to OpenSSL’s base64 encoder
  • And send the result of all of this to pbcopy (you can use clip on Windows)

The above produced the following:


We then use the above as the value for our integrity attribute as follows:


With the above in place, your browser can verify that the contents received from the CDN are the contents you expect and only parse/execute/apply the code to the current context if verifications pass

You can use the same mechanism with preload and modulepreload which we discussed previously.

That is it! πŸ’₯ You now have the means to securely load your content from a third party πŸŽ‰

Buy me a coffee and support the work on 150 Days of HTML.