Re: alternative to _XLATEB on Windows

Well, coding up a single-byte-to-single-byte bit of code (at least if you
aren't fussy about performance) is about a five minute job.  You then just
need the tables.  Dealing with Asian character sets is a much bigger job.

Consult your Microsoft (or Borland?) C/C++ manual (see "Character sets").

I only have MS C++ Version 5 and the internationalization support there is
pretty weak.  But, to be fair, pretty old.

Probably a long shot on a Windows system, but you might see if the iconv
interface has been made available somehow.  This is available in some form
on most IBM systems and is part of the Single Unix specification.  But,
apparently, Microsoft has reinvented the wheel, here.  Or, at least it had
as of MS C++ 5.  Microsoft seemed then to think people only need to
translate to the local character set and/or to Unicode, or at least it
thought that way several years ago.

IBM used to sell a CD ROM with extensive character set documentation
including, I believe, tables and even some conversion code.  Search on CDRA
(I think it is) and see what turns up.

Finally, Java (and IBM's Toolbox for Java) has a fair amount of this
built-in one way or another.  Java's Reader and Writer classes do
conversion to and from Unicode that can be implicitly changed by setting a
Java property.  The Toolbox has a class named either Text or AS400Text
(forget which) that I think had good conversions in it.  Since many
practical transations only exist with Unicode as an intermediary (there's
too many character sets in the world to translate all to all), the Java
unicode-as-intermediary can be a good way to go in any event.



Larry W. Loen  -   Senior Linux, Java, and iSeries Performance Analyst
                          Dept HP4, Rochester MN