Frequently Asked Questions
Non-ASCII characters are missing or incorrect
Q.
Some characters are replaced with question marks on my pages
A.
This is most commonly caused by:
- A missing DOCTYPE in your page template
- A missing or incorrect character set in your page template
- Using characters from a different character set (e.g. latin1 characters, when the charset is utf-8)
- The
mbstring.encoding_translationsetting is turned on. This automatically converts all page output to some other character set, often ISO8859-1.
You can check the mbstring settings with php -i or phpinfo();. If mbstring.encoding_translation is on, ask your host about turning it off.
Textpattern has built-in support for Unicode. This means you don’t have to do anything special for non-ASCII characters to work. Just write articles using a modern browser, and enter any characters you like in the article title, body, and excerpt.
Don’t specify a legacy character set in your page doctype or HTML <meta> tags. The character set for all Textpattern pages should always be “utf-8”. If you use anything else, non-ASCII characters will be encoded incorrectly.
Posted 17 August 2005, 10:26 by Alex Shiels
