CodeIgniter config charset and UTF-8 support

Question

I'm working on a website which uses / stores accented characters in the database. I have the page template set so that the config.php charset variable matches the setting, e.g.:

<meta charset="<?php echo $this->config->item('charset');?>">

The problem I'm having is, when $config['charset'] is set to UTF-8, the form validation fails and it's as if no characters were submitted if an accented character was included. So, for example, a required field will bounce back if á is included anywhere in the string. The string minus the á works fine.

I've managed to get this working by changing the $config['charset'] to ISO-8859-1 and converting text to UTF-8 before inserting / after retrieving from the database with php's utf8_encode() and utf8_decode(). Is this the best way or am I missing something needed in order to get UTF-8, with accented characters, working in CodeIgniter?

Any advice appreciated.

Solution

You have to make sure that you use UTF-8 everywhere, and that both PHP and MySQL are configure to handle UTF-8.

In the html, add the meta-tag:

<meta charset="utf-8" />

And save it in UTF-8 format. here is how to do that in notepad++.

Define the MySQL tables to support UTF-8, create table with:

DEFAULT CHARSET=utf8;

And set the connection to:

mysql_set_charset('utf8', $con);

Enable UTF-8 in the php.ini:

default_charset = "utf-8"

For a full manual check Handling Unicode Front To Back In A Web App

Comments

[Dan Murfitt]

Many thanks for the answer and link - I'll check to make sure UTF-8 is being used throughout. The form validation problem was being caused by the $config['charset'] setting being set to 'UTF-8' (all uppercase) rather than 'utf-8' (all lowercase). For some reason the default setting for this being uppercase 'UTF-8' doesn't work (at least not on my setup).

[MarcinWolny]

Small comment for people who found it on google - alternative source of this problem might be encoding of the table or column (yes, changing encoding of the table does not change the encoding of individual columns) - in that case editing value within a column with UTF-8 string would result in "Incorrect string value” warning, which is not visible when you simply run query through codeigniter. All you'll see is ? instead of UTF-8 char.