329

How can I strip the HTML from a string in JavaScript?

1
  • result = text?.replace(/<\/?[^>]+(>|$)/g, ""); Commented Oct 1, 2024 at 13:12

4 Answers 4

521
cleanText = strInputCode.replace(/<\/?[^>]+(>|$)/g, "");

Distilled from this website (web.achive).

This regex looks for <, an optional slash /, one or more characters that are not >, then either > or $ (the end of the line)

Examples:

'<div>Hello</div>' ==> 'Hello'
 ^^^^^     ^^^^^^
'Unterminated Tag <b' ==> 'Unterminated Tag '
                  ^^

But it is not bulletproof:

'If you are < 13 you cannot register' ==> 'If you are '
            ^^^^^^^^^^^^^^^^^^^^^^^^
'<div data="score > 42">Hello</div>' ==> ' 42">Hello'
 ^^^^^^^^^^^^^^^^^^          ^^^^^^

If someone is trying to break your application, this regex will not protect you. It should only be used if you already know the format of your input. As other knowledgable and mostly sane people have pointed out, to safely strip tags, you must use a parser.

If you do not have access to a convenient parser like the DOM, and you cannot trust your input to be in the right format, you may be better off using a package like sanitize-html, and also other sanitizers are available.

Sign up to request clarification or add additional context in comments.

14 Comments

Sorry, but that would break <img alt="a>b" src="a_b.gif" />
@f.ardelian people who make a hobby out of breaking the ill-use of regular expressions for parsing general HTML are great. It is a great hobby.
@f.ardelian no! Really! Every time I read one of these comment threads I get a little thrill. "Ho ho ho," I think "<img alt=\"a>b\" src=\"a_b.gif\" />, so clever!"
@f.ardelian That would be buggy html, it had to be <img alt="a&gt;b" .
using reg is not good approach stackoverflow.com/questions/1732348/…
|
330

Using the browser's parser is the probably the best bet in current browsers. The following will work, with the following caveats:

  • Your HTML is valid within a <div> element. HTML contained within <body> or <html> or <head> tags is not valid within a <div> and may therefore not be parsed correctly.
  • textContent (the DOM standard property) and innerText (non-standard) properties are not identical. For example, textContent will include text within a <script> element while innerText will not (in most browsers). This only affects IE <=8, which is the only major browser not to support textContent.
  • The HTML does not contain <script> elements.
  • The HTML is not null
  • The HTML comes from a trusted source. Using this with arbitrary HTML allows arbitrary untrusted JavaScript to be executed. This example is from a comment by Mike Samuel on the duplicate question: <img onerror='alert(\"could run arbitrary JS here\")' src=bogus>

Code:

var html = "<p>Some HTML</p>";
var div = document.createElement("div");
div.innerHTML = html;
var text = div.textContent || div.innerText || "";

14 Comments

Doesn't work when the string contains something like <script>alert('hi');</script>. Then it crashes with "illegal token at" etc..
Good caveats. In case it is not already clear I wanted to add that Firefox will crash on div.innerHTML = html if the value of html is NULL. Worse, it won't properly report the error (instead says parent function has TypeError). Chrome/IE do not crash.
SECURITY ISSUE ... This could be vulnerable as you're setting div.innerHTML ... i'm sure you don't wanted to get some unwanted script executed. ... manual cleanup would be cool.
Elegant solution, but isn't universal. It doesn't work if you use it on node server because of the document dependency
<p>test</p><p>test</p> gives testtest, should have spave or newline between
|
93
var html = "<p>Hello, <b>World</b>";
var div = document.createElement("div");
div.innerHTML = html;
alert(div.innerText); // Hello, World

That pretty much the best way of doing it, you're letting the browser do what it does best -- parse HTML.


Edit: As noted in the comments below, this is not the most cross-browser solution. The most cross-browser solution would be to recursively go through all the children of the element and concatenate all text nodes that you find. However, if you're using jQuery, it already does it for you:

alert($("<p>Hello, <b>World</b></p>").text());

Check out the text method.

10 Comments

Not every browser supports innerText.
A concise jQuery could look like: var html = "<b>test</b>"; var text = $("<div/>").html(html).text(); Using $("<div/>") lets you reuse the same element and less memory for consecutive calls or for loops.
Same problem, crash it with: $(...).html('<script>alert("hi");</script>').text();
and check out the text method for var txt = "<p>my line</p><p>my other line</p>some other text"; alert($(txt).text(); where you don't proxy the string within a dom node. 3 lines in, 2 lines out.
The jQuery solution is the best for all of us (most of us, I guess) who already use it almost everywhere. Just keep in mind that if the string is in a variable, you will have to insert it in an element, e.g. let text = $(`<div>${html_fragment}</div>`).
|
36

I know this question has an accepted answer, but I feel that it doesn't work in all cases.

For completeness and since I spent too much time on this, here is what we did: we ended up using a function from php.js (which is a pretty nice library for those more familiar with PHP but also doing a little JavaScript every now and then):

http://phpjs.org/functions/strip_tags:535

It seemed to be the only piece of JavaScript code which successfully dealt with all the different kinds of input I stuffed into my application. That is, without breaking it – see my comments about the <script /> tag above.

4 Comments

^ this, definitely better than the accepted answer for Chrome 30.0 and above
Works nicely on server-side without DOM support, e.g. Google Apps Script.
If you use the allowed param you are vulnerable to XSS: stripTags('<p onclick="alert(1)">mytext</p>', '<p>') returns <p onclick="alert(1)">mytext</p>

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.