On some occasions I use to code html pages with Romanian special characters. Unlike with the German special characters (ä etc.) I feel better to use the standardized entities for Romanian characters in the form ă etc.
In TextMate I could not find out yet how to convert a text selection with Romanian special characters typed so that only the special characters are converted to entities. In my page *every* single character is replaced by its entity and makes the text illegible.
Where is the solution I missed in the manual?