How to detect the right font to use depending on the langage

For a program of mine I have a database full of street name (using GIS stuff) in unicode. The user selects any part of the world he wants to see (using openstreetmap, google maps or whatever) and my program displays every streets selected using a nice font to show their names. As you may know not every font can display non latin characters... and it gives me headaches. I wonder how to tell my program "if this word is written in chinese, then use a chinese font".

EDIT: I forgot to mention that I want to use non-standard fonts. Arial, Courier and some other can display non-latin words, but I want to use other fonts (I have a specific font for chinese, another one for japanese, another one for arabic...). I just have to开发者_StackOverflow中文版 know what font to chose depending of the word I want to write.

You need information about the language of the text. And when you decide what fonts you want, you do a mapping from language to font.

If you try to do it automatically, it does not work. The fonts for Japanese, Chinese Traditional, and Chinese Simplified look differently even for the same character. They might be inteligible, but a native would be able to tell (ok, complain) that the font is wrong.

Plus, if you do anything algorithmically, there is no way to consider the estethic part (for instance the fact that you don't like Arial :-)

Use utf-8 text and a font that has glyphs for every possible character defined, like Arial/Verdana in Windows. That bypasses the entire detection problem. One font will handle everything.