竹貓星球

12. I am seeing &#nnnn; sequences output instead of what I typed!

To enable compliance with HTML and XHTML specifications as well as enabling browsers to automatically support a users selected language we enable a charset tag to be output to the browser. This identifies the primary character set the user has selected, eg. iso-8859-1 for Western European languages (English, German, French, etc.), windows-1251 for Cyrillic alphabets (Russian, Bulgarian, etc.) and so on. The downside to doing this is that most modern browsers (such as Internet Explorer) convert any character not present in that character set into what are termed Numerical Character References or NCR's. These take the form of &#nnnn; where nnnn is the hexadecimal number of that characters UNICODE reference.

Unfortunately PHP (which effectively supports only iso-8859-1 at this time) does not appear to attempt to process this data back into normal characters. Therefore phpBB treats them as text the user has entered and wishes to see. phpBB enables this by turning the & into &. This results in the characters appearing as they do. Although we could ignore &#nnnn; (there are pro's and of course con's to doing so) at present we don't.

If you are finding this to be problem the solution is to remove the

代碼: 選擇全部

charset={S_CONTENT_ENCODING}

text from all the _header.tpl template files in all templates. The downside to this is a users browser will not automatically switch its charset to that selected by the user (however this is unlikely to be a big problem since users typically run their browser by default in the language they wish to use). Alternatively you can modify the $lang['ENCODING'] entry in the appropriate language file. Until PHP (and all browsers!) trully support UNICODE these problems will persist across all boards.