How to substr html entities properly?

I have like this :

$mytext="that's really "confusing" and <absolutly> silly";
echo substr($mytext,0,6);

开发者_运维知识库

The output in this case will be : that&# instead of that's

What i want is to count html entities as 1 character then substr, because i always end up with breaked html or some obscure characters at the end of text.

Please don't suggest me to html decode it then substr then encode it, i want a clean method :)

Thanks

---

There are two ways of doing this:

1. You can decode the HTML entities, substr() and then encode; or

2. You can use a regular expression.

(1) uses html_entity_decode() and htmlentities():

$s = html_entity_decode($mytext);
$sub = substr($s, 0, 6);
echo htmlentities($sub);

(2) might be something like:

if (preg_match('!^([^&]|&(?:.*?;)){0,5}!s', $mytext, $match)) {
  echo $match[0];
}

What this is saying is: find me up to 5 occurrences of the preceding expression from the beginning of the string. The preceding expression is either:

• any character that isn't an ampersand; or

• an ampersand, followed by anything up to and including a semi-colon (ie an HTML entity).

This isn't perfect so I would favour (1).

---

function encoded_substr($string, $param, $param2){
  $s = html_entity_decode($string);
  $sub = substr($s, $param, $param2);
  return htmlentities($sub);
}

There, I copypasted cletus' code into a function for you. Now you can call a very straightforward 3 line function with 1 line of code. If this isn't "clean" then I'm confused what "clean" means.

---

Be advised that some characters break the proposed decoding + encoding, if you use substr().

Example

$string=html_entity_decode("Workin’ on my Fitness…In the Backyard.");

echo $string;
echo substr($string,0,25);
echo htmlentities(substr($string,0,25));

Will output:

• Workin’ on my Fitness…In the Backyard.
• Workin’ on my Fitness�
• (empty string)

The solution

Use mb_substr().

echo mb_substr($string,0,25);
echo htmlentities(mb_substr($string,0,25));

Will output:

• Workin’ on my Fitness…In
• Workin’ on my Fitness…In

---

Please try with following coding Functions.

<?php

$mytext="that&#039;s really &quot;confusing&quot; and &lt;absolutly&gt; silly";

echo limit_text($tamil_var,6);

function limit_text($text,$limit){
   preg_match_all("/&(.*)\;/U", $text, $pat_array);
   $additional=0;

   foreach ($pat_array[0] as $key => $value) {
     if($key <$limit){$additional += (strlen($value)-1);}
   }
   $limit+=$additional;

   if(strlen($text)>$limit){
     $text = substr( $text,0,$limit );
     $text = substr( $text,0,-(strlen(strrchr($text,' '))) );
   }
   return $text;

}

?>

---

Well, clean method is only one: Not to use entities at all.
There are not a single reason to substr entitied string. It can be used to output only.
So, first substr, then encode.

---

Here is a correction for syntax error code, use mb_substr to avoid surprises like html entity having less characters, or character counting not working the way it should, in my case Sábado becoming Sá:

function encoded_substr($string, $param, $param2){
$s = html_entity_decode($string);
$sub = mb_substr($s, $param, $param2);
return htmlentities($sub);
}