How to truncate an UTF8 string in PHP?

Question

I have some strings in my PHP code that need to be truncated if they are too long.

For example if a text is something like this:

Hi, I would like to tell you how wonderful this is.

It would replace it with this:

Hi, I would like to ...

For that I've done a simple substr. The problem is that in UTF8 some characters are actually two characters long. And I've had some problems with a character being cut in the middle: For example, when I try to insert the modified string in the database, it crashes.

Here is my current function:

static function short($string, $max = 255){
   if(strlen($string) >= $max){
       $string = substr($string, 0, $max - 5).'...';
   } return $string;
}

Would someone know a way to make this function work even for UTF8 characters?

Solution 1

Everything you need is mb_strimwidth() : http://php.net/manual/en/function.mb-strimwidth.php

Example:

mb_strimwidth('Hi, I would like to tell you how wonderful this is.',0,15,'...','utf-8');

Solution 2

try with mb_substr() :

static function short($string, $max = 255){
   if(mb_strlen($string, 'utf-8') >= $max){
       $string = mb_substr($string, 0, $max - 5, 'utf-8').'...';
   } return $string;
}

Comments

[caw]

You might find s($str)->truncate($length) and s($str)->truncateSafely($length) helpful, as found in this standalone library. Both versions are Unicode-safe. The latter does not break words.

[Tudor Constantin]

@deceze - you are right, thx for pointing that out - i corrected the copy/pasted code

[Álvaro González]

@trejder No, it isn't. The initial version of this article was claiming that mb_strimwidth() is deprecated in PHP 7.2 and that this deprecation was a source of described problem. It turned out that mb_strimwidth() is (of course!) not deprecated in PHP 7.2.

[trejder]

@ÁlvaroGonzález Correct! I'm removing my comment. Thanks!