This guide is intended for developers and webmasters who are experiencing display issues with special characters (e.g., accents, Arabic characters, Chinese characters, emojis, etc.) on a website or in a database hosted by Infomaniak.

Introduction

• Common causes of incorrectly displayed characters that appear instead of special characters or emojis like ??:
  • PHP connection in utf8 instead of utf8mb4
  • MySQL table or column in latin1 or utf8 instead of utf8mb4
  • SQL file encoded differently from the import setting
  • Incorrectly encoded HTML/CSS/JS files
  • Missing or incorrect charset declaration on the HTML, PHP, or HTTP side
• Common solutions:
  • Use utf8mb4 everywhere: database, connections, tables, columns, files, headers
  • Validate the encoding of all files and dumps
  • Test the display of emojis, accents, and other multilingual characters
• Also, refer to this other guide to force a different encoding when connecting to a MySQL database.

1. Use UTF-8 or UTF-8MB4 everywhere

MySQL's utf8 encoding does not support 4-byte characters, such as emojis. These will be displayed as ??. Therefore, you should use utf8mb4, which is a true, complete implementation of UTF-8.

2. HTML – Correct encoding declaration

In the <head> of your HTML pages:

<meta charset="UTF-8">

Also, make sure that the CSS/JS files are saved in UTF-8 (without BOM).

header('Content-Type: text/html; charset=utf-8');

$connection = mysql_connect($host, $user, $pass);
mysql_set_charset('utf8mb4', $connection);

 $dsn = "mysql:host=localhost;dbname=ma_base;charset=utf8mb4";
$pdo = new PDO($dsn, $user, $pass, [
    PDO::ATTR_ERRMODE => PDO::ERRMODE_EXCEPTION
]);

ALTER DATABASE ma_base CHARACTER SET = utf8mb4 COLLATE = utf8mb4_unicode_ci;
ALTER TABLE ma_table CONVERT TO CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;

mysqli_query($conn, "SET NAMES 'utf8mb4'");
mysqli_query($conn, "SET CHARACTER SET 'utf8mb4'");

AddDefaultCharset utf-8
Header set Content-Type "text/html; charset=utf-8"
# Pour PHP
php_value default_charset UTF-8
php_value mbstring.internal_encoding UTF-8

iconv -f ISO-8859-1 -t UTF-8 dump.sql > dump_utf8.sql