How to calculate byte length containing UTF8 characters using javascript?

Question

I have textbox, in which the user can enter the characters in ASCII/UTF-8 or a combination of both. Is there any API in javascript which we can calculate the length of string in bytes for the characters entered in textbox.

Like if i enter ascii chacter let's say : mystring - the length would be calculated as 8. But when UTF8 characters are entered the characters can be 2/3/4 byte.

lets say the character entered : i ♥ u , the length in bytes is 5.

The textbox can accept max length of 31 characters. But in case if UTF8 characters entered, it will not accept character string : i ♥ u i ♥ u i ♥ u i ♥ u i ♥ u . the length is 30.

Can we restrict the user to enter characters not more than 31 even for UTF8 characters.

Answer 1

Counting UTF8 bytes comes up quite a bit in JavaScript, a bit of looking around and you'll find a number of libraries (here's one example: https://github.com/mathiasbynens/utf8.js) that can help. I also found a thread (https://gist.github.com/mathiasbynens/1010324) full of solutions specifically for utf8 byte counts.

Here is the smallest, and most accurate function out of that thread:

function countUtf8Bytes(s){
    var b = 0, i = 0, c
    for(;c=s.charCodeAt(i++);b+=c>>11?3:c>>7?2:1);
    return b
}

Note: I rearranged it a bit so that the signature is easier to read. However its still a very compact function that might be hard to understand for some.

You can check its results with this tool: https://mothereff.in/byte-counter

One correction to your OP, the example string you provided i ♥ u is actually 7 bytes, this function does count it correctly.

Answer 2

As of 2018, the most compatible and reliable way of doing this seems to be with the blob api.

new Blob([str]).size

Even supported in IE10 if anyone uses that anymore.

Answer 3

The experimental TextEncoder API can be used for this but is not supported by Internet Explorer or Safari:

(new TextEncoder()).encode("i ♥ u i ♥ u i ♥ u i ♥ u i ♥ u").length;

Another alternative is to URI-encode the string and count characters and %-encoded escape sequences, as in this library:

~-encodeURI("i ♥ u i ♥ u i ♥ u i ♥ u i ♥ u").split(/%..|./).length

The github page has a compatibility list which unfortunately does not include IE10, but IE9.

Since I can not yet comment, I'll also note here that the solution in the accepted answer does not work for code points consisting of multiple UTF-16 code units.