I’ve got a head scratcher that I’m not sure how to deal with. I’m using htmlspecialchars to display user inputted variables on a page. On my local server (PHP version 5.1.4) it works just as expected, but on the live site (PHP version 5.2.9), if there are any accented characters in the string, I am getting a null value after running it through the htmlspecialchars function.
Here is some sample code I’ve been working with, if it’s at all helpful.
$text = "canapés";
echo 'My text is '. $text;
$text = htmlspecialchars($text, ENT_QUOTES, 'UTF-8');
echo 'The encoded text is ' . $text;
I’m unable to reproduce your problem (PHP 5.3, Apache 2, Windows). However, you shouldn’t need to specify the charset, since, in your case, the characters affected by the function are in the same positions as in ISO-8859-1 (htmlspecialchars() docs).
Yep, that was the whole code that I was using for testing. After the first suggestion to use htmlentities instead (which works fine) I started poking at it some more, and what I’ve found is that if I don’t specify the charset, then it works.
$text = 'canapés aren\\'t popular';
echo 'My text is '. $text;
$text = stripslashes(htmlentities($text, ENT_QUOTES));
echo 'The htmlentities text is ' . $text;
$text = 'canapés aren\\'t popular';
echo 'My text is '. $text;
$text = htmlspecialchars($text, ENT_QUOTES);
$text=str_replace('&','&',$text);
echo 'The encoded text is ' . $text;
Output:
My text is canapés aren't popular
The html entities text is canapés aren& #039;t popular
My text is canapés aren't popular
The encoded text is canapés aren&# 039;t popular
I’m the first to admit that what I know about character encoding would fit on the head of a pin, but I’m particularly concerned about it for this project because the site serves an international audience and needs to be able to accurately reproduce character sets from multiple languages.
P.S. No matter what I do, the forum software is converting my apostrophes so I stuck spaces in there to try to force it.
I’m afraid I’m at a loss, since I’m still unable to reproduce the problem. Just out of curiosity, though, what is the output on your live server for the following:
I cant really explain it either, since neither of the hex-pairs that constitute e-acute are a quote character… perhaps it’s being screwed up by the character being read as e’ or e` or some other Quote-Containing phrase? Have you tried parsing the string without the ENT_QUOTE flag?
So, it appears the accented character is not being interpreted properly. Is it possible the server itself can’t handle UTF-8 encoding? Do character sets have to be enabled somehow on a server?
Great, can you throw the following in a stand-alone script and post both the headers and output?
<?php
header('Content-Type: text/plain; charset=UTF-8');
echo htmlspecialchars("Anthony's canapés aren't popular at all, in fact, they suck.", ENT_QUOTES, 'UTF-8');
exit;
<?php
header('Content-Type: text/plain; charset=UTF-8');
echo htmlspecialchars(
utf8_encode("Anthony's canapés aren't popular at all, in fact, they suck."),
ENT_QUOTES | ENT_COMPAT,
'UTF-8'
);
exit;
If it works, your PHP script is saved by your editor as ISO-8859-1.
Not quite The forum software converted my apostrophes. The output doesn’t convert the encoding for the apostrophes, so they come out as & #039; If I change the header, is that a bad thing to do?
The forum software converted my apostrophes. The output doesn’t convert the encoding for the apostrophes, so they come out as &