I have made web-app that uses iso-latin-2 charset. I'll describe shortly what happens with my application. ----------------------------
I set meta tag charset=iso-8859-2 on every .jsp page.
When I receive data from my input form, some characters (that are out of iso-8859-1, i.e. belongs to iso-8859-2) comes in form of &#nnn; (e.g. ?=š č=č etc.) That is OK!!!
For example, if an input in a textfield (named "school") on the .jsp page that contains form was: ?kola after pressing Submit, on the following .jsp page I can read that parameter as:
String str = request.getParameter("school");
Parameter str is now equal to škola
When I print parameter "str" onto my .jsp page it appears OK:
<%= str %> produces ?kola when rendered onto .jsp page
Storing data that contain that sort of chars (e.g. š, č ć etc.) into my database, makes it longer than it should be, e.g.:
?kola - from my input form (length=5) is equal to: &#353;kola - when stored into database (length=10)
I want to write into my database datas of exactly the same size as they apper when rendered in an HTML(JSP) page.