certain characters behaving unusually within htmlentities function

Question

So first I noticed when an ampersand (&) is in the htmlentities function it will count the ampersand as 5 characters. So this code:

$a = htmlentities("&12345");
$b = substr($a,0,6);
  echo $b; 

would echo '&1' as, I believe it is counting & as 5 characters. more interestingly is with the GBP symbol (£) which is ignored all together so this:

$a = htmlentities("£");
  echo $a;   

echos nothing. I get the same results on Chrome and FF. I don't know if this is a bug or I should use a different syntax. Does anyone know why this might be happening? Thanks

UPDATE

I have solved the £ issue with this: $a = htmlentities("£", ENT_COMPAT, 'ISO-8859-15'); but the ampersand issue remains.

Answer 1

The problem is that you are viewing the output in the browser window where output from htmlentities (and equivalent functions) is rendered for display to an end user.

For example:

echo htmlentities("&");

Will output & because it has converted the & character to it's html entity equivalent. See that it is five characters long. However, you don't see the full text & because you're viewing from your browser which has pre-rendered it as the & symbol. In firfox if you right click the viewport and click on "View Page Source" you will see the full text &...

Your code:

$a = htmlentities("&12345"); //Outputs: &12345
$b = substr($a, 0, 6); //Selects first six charachters: &, a, m, p, ; and 1
echo $b; // Echo's: &1 which is displayed by the browser as &1

To get around this you can change the order of your functions:

$a = substr("&12345", 0, 6);
echo htmlentities($a);