U++ forum: Welcome to the forum

Re: How to display several languages without translation files? [message #4443 is a reply to message #4441]

Tue, 08 August 2006 20:29

mirek
Messages: 13976
Registered: November 2005

Ultimate Member

Beware, scope of SetDefaultCharset is the whole application (I cannot imagine how to limit its scope to package only....)

I guess this needs more explainint.

DefaultCharset is single global variable. Value of this variable is used on a couple of absolutely critical places to convert char to wchar and back when no explicit charset (codepage) is specified, namely:

String <-> WString
host OS text (e.g. filesystem names)
text translations via t_ function (all translations are stored in UTF-8, t_ translates them to current default charset).

Maybe I forgot something, but these three points are mostly enough to have your app in any encoding you need.

Speaking about UTF-8, note that U++ UTF-8 has a little "improvement" - standard UTF-8 has problem as not every combination of bytes is valid UTF-8 text (means you can have "UTF-8 format errors").

This causes stupid problem for UTF-8 capable editors (e.g. the source editor in TheIDE) - if you have a file that contains multiple encodings (why not...), you cannot open it in the editor.

Therefore, U++ extends UTF-8 to cover these cases by encoding "errors" using "Unicode private area" - 0xee00 - 0xeeff. This way, you can load and save such files without problems. (Nice name for this format is therefore UTF-8EE - could mean "error extension" and is the same as hex address Wink

Mirek

Report message to a moderator