Coding system

[mara]

Hi every body,

Am willing to write a foreighn language driver for my os, does any one has an idea about the coding system I might use and how to impelment it ?

[smiddy]

Yupper, gather up all you output text into one table. Use the table to get the text output for the particular instance it is needed. Then use a translated table with the same offsets for the other lanaguages, you could do any number of languages this way and make it interchangeable too.

[matthias]

Yupper, gather up all you output text into one table. Use the table to get the text output for the particular instance it is needed. Then use a translated table with the same offsets for the other lanaguages, you could do any number of languages this way and make it interchangeable too.

Wow... good idea

[mara]

Thank you for help, but it's a deeper issue!
The problem is with the "Coding system itself", the method mentionned above is very handy especialy with one byte coding ("flat" english, dos oem...) but if the UTF-8 or UTF-16 is chosen???

There will be endianism problems and code detection ones...

[smiddy]

In the case of different ENDIANs then you would establish a database structure with the particular information for the particular language and output style. Considering that some languages have thousands of characters it would be useful to set the database up with specific fields for given languages and use the appropriate system for the given language. You could use a dBase III style setup (real easy to learn and implement) or you could create your own, a type of compression scheme since not all items are the same length. What ever you do though a look up table of sorts will be needed. There is also a limit to the number of languages so some analysis will have to go into similarities and figure out algorythms appropriate for each given one. This is extensible you know, you can do it. I would be interested in your own thoughts...