Published at
Updated at
Reading time

This post is part of my Today I learned series in which I share all my learnings regarding web development.

Today I came across a blog post (it's in German though) written by Peter Kröner and learned something very astonishing.

The article describes the not so well known behaviors of the method String.prototype.replace. Using this method is usually very straight forward. It supports regular expressions if you need to, but for most cases, it's defining a matching string and a replacing string.

That's at least what I thought... 🙈

It turns out that if the second argument is a string (it can also be a function) and includes specific character sequences like $& or $' "replacer magic" appears.

const msg = 'This is a great message';

msg.replace('great', 'wonderful'); 
// "This is a wonderful message"
// -> 'great' is replaced by 'wonderful'

msg.replace('great', '$&-$&');
// "This is a great-great message"
// '$&' represents the matched substring
// -> 'great' is replaced by 'great-great'

msg.replace('great', '$`');
// "This is a This is a  message"
// '$`' represents the string before the matched string
// -> 'great' is replaced by 'This is a '

msg.replace('great', `$'`)
// "This is a  message message"
// `$'` represents the string after the matched string
// -> 'great' is replaced by ' message'

Oh my..., this behavior can lead to very hard to spot bugs!

If you want to read more about it, have a look at the replace docs on MDN or this overview on replacement values.

Related Topics

Related Articles