PHP 5.6 default_charset change may break HTML output

An important note for everyone who’s upgrading from PHP 5.4 and PHP 5.5, to PHP 5.6: the PHP default_charset in php.ini changed from “empty” to UTF-8, making UTF-8 the default charset in PHP. This may break HTML output if you try to set a different charset in your HTML head. It may also break functions like htmlentities() and htmlspecialchars. For example:

Breaking HTML output with PHP 5.6 default_charset changed to UTF-8

Suppose you have the following lines in PHP code (don’t ask me why you would want to do this…):

<html> 
<head> 
<meta http-equiv="content-type"
  content="text/html; charset=ISO-8859-1"> 
</head>
<body> 
<?php echo "éééééeeeeééé"; ?>
</body> 
</html>

The PHP default charset setting in PHP 5.4 and PHP 5.5 prints the expected string on the screen: éééééeeeeééé.

In PHP 5.6 however, the default_charset is set to UTF-8, changing PHP’s default charset, and PHP 5.6 will always print a Content-Type response header set to UTF-8:

Content-Type: text/html; charset=UTF-8

The PHP default charset set to UTF-8 breaks HTML output and functions like htmlentities() / htmlspecialchars() (PHP bug #61354), because the HTML charset ISO-8859-1 is printed too (creating a double Content-Type response header):

GET -uUsSed http://www.example.com/echo.php
GET http://www.example.com/echo.php
User-Agent: lwp-request/2.07

GET http://www.example.com/echo.php --> 200 OK
Cache-Control: private
Date: Tue, 31 Mar 2015 12:03:50 GMT
Server: Microsoft-IIS/8.0
Content-Length: 141
Content-Type: text/html; charset=UTF-8
Content-Type: text/html; charset=ISO-8859-1
Client-Date: Tue, 31 Mar 2015 12:03:50 GMT
Client-Peer: 77.94.251.244:80
Client-Response-Num: 1

Pro Tip: While going through your PHP config, fix & set correct values for curl.cainfo and openssl.cafile too. Don’t turn off CURLOPT_SSL_VERIFYPEER (or sslverify = false for wp_remote_get() in WordPress).

Interesting:   "htaccess files should not be used for security restrictions"

PHP.net writes:

If omitted, the default value of the encoding varies depending on the PHP version in use. In PHP 5.6 and later, the default_charset configuration option is used as the default value. PHP 5.4 and 5.5 will use UTF-8 as the default. Earlier versions of PHP use ISO-8859-1.

Although this argument is technically optional, you are highly encouraged to specify the correct value for your code if you are using PHP 5.5 or earlier, or if your default_charset configuration option may be set incorrectly for the given input.

How to fix the HTML output in PHP 5.6 – and up (7.0, 7.1)

The most obvious solution to this problem is: don’t set a character set encoding in your HTML meta tag, e.g:

<meta http-equiv="content-type"
  content="text/html; charset=ISO-8859-1">

Other options to set a correct default character set (default_charset) are:

Create an user-defined php.ini to overrule default_charset directive

PHP supports user-defined php.ini files, in which you can overrule some php.ini settings. Neat!

Upload your user-defined php.ini to your webroot containing the following line:

default_charset = ""

This will tell PHP to not send a Content-Type response header set to UTF-8.

Overrule default_charset with ini_set()

And last but not least, you can overrule this setting with PHP’s ini_set() function:

ini_set( default_charset, "" );

Please share this post if you found it useful, thank you! If you have a valuable tip, please let me know and drop me a comment.

4 replies
  1. RemBem
    RemBem says:

    Thank you, this saved my day, fixing an old site suddenly full of questionmarks after upgrading php. All your other website performance articles are also very helpful!




    1



    0
    Reply
  2. Anonymous
    Anonymous says:

    In PHP 7.0.19, the empty value of the default_charset directive (default_charset = “”) causes an error HTTP 500 “Internal Server Error”.
    You can solve this problem by setting the value to “none”. For example:
    php_value default_charset = none
    or
    php_value default_charset = “none”
    or
    php_value default_charset = ‘none’




    0



    0
    Reply

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published.