Hi there,

I'm writing a program in C++ with wxWidgets that has to output characters in different scripts e.g. Devanagari, Tibetan etc. When I need to output a character that is made up of two characters (by making a string like "\u0F63\u0FA1") they don't come out stacked in the correct way, (an analogy with latin characters would be é and É where the accent sign doesn't move for E), I looked in the font I'm using and it has a separate glyph for the two chracters stacked but it does not get ouput, instead \u0F63 is output and then \u0FA1 is output underneath, so they are "semi stacked".

Is it my responsibility as the programmer to look at the two characters and select the apropriate glyph from the font, or should it be handled by the operating system/X server/desktop enviroment and something is not configured properly?

Also, if anyone could point me in the direction of some good resources unicode and C++ it would be much appreciated.

Thanks

This should be managed by the OS, first check that this works on your system. Use any Unicode aware application and check the script. If that works then your OS is probably fine but the glyph rendering engine that your program is using (via wxWidgets) is outdated. I say probably because there are a few rendering engines, I forget their names.

For Tibetan check out the Pan Localisation project (Dzongkha). I was involved in this in the very early days and I know that they have got it working but you may require an update to your system.

Thanks for that information. Firefox displays Tibetan letters fine and so do all the Java programs that use Tibetan I've tried, so I will look into updating my system.

Can anyone recommend any websites or books that have good information about localization, Unicode, etc.? Because I've found it had to find information on supporting languages other than English in my programs.

Thanks

Getting info on programming Unicode is not easy and since C/C++ doesn't support it natively you will need to look for various libraries. What I can suggest is that you look at the PANL10N web site and the Dzongkha project on Source forge. There are some contact emails you can extract from those sites to ask people about specific questions that you have.

By the way Dzongkha is the language from Bhutan and it uses the Tibetan script.

Check out Markus Kuhn's indispensable UTF-8 and Unicode FAQ. A must-read.

Also, O'Reilly's old "CJKV" book is sometimes helpful, but they've got a new one actually called "Unicode explained" and another one "Fonts and Encoding". I don't know them, though.

Thanks for all your help, those resources should be really helpful.