Hmmm... So you are saying its resource compiler is misinterpreting the text
in the RC file? That would be problematic. If both the VC++ resource
compiler (under both NT and 98) and the Borland resource compiler would
directly take the L"abcde" style wide character constants, instead of the
escaped versions, then I'd be glad to change it to use that style. But I'm
kinda thinking that perhaps 98 wouldn't be able to eat those.
Perhaps you can figure out how to make Borland take wide chars in the RC
file? Since the RC file format is Unicode anyway, its difficult to imagine
that the resource compiler can't understand Unicode.
----------------------------------------
Dean Roddey
Software Weenie
IBM Center for Java Technology - Silicon Valley
[EMAIL PROTECTED]
Mike Goodey <[EMAIL PROTECTED]>@compuserve.com> on 04/05/2000
05:36:15 AM
Please respond to [EMAIL PROTECTED]
To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>, Dean
Roddey/Cupertino/[EMAIL PROTECTED]
cc:
Subject: Re: xerces c and Borland
Dean
I'm getting a bit further with this. And I'm using NT 4.
What is happening is that the resource lines are being interpreted
differently by Borland
e.g. The line (well word wrapped by now)
2
L"\x004E\x006F\x0074\x0061\x0074\x0069\x006F\x006E\x0020\x0027\x007B\x0
030\x007D\x0027\x0020\x0068\x0061\x0073\x0020\x0061\x006C\x0072\x0065\x
0061\x0064\x0079\x0020\x0062\x0065\x0065\x006E\x0020\x0064\x0065\x0063\
x006C\x0061\x0072\x0065\x0064\x00"
is interpreted as the (wide) string
L"\004e\006f...
where \0 is a binary 0, and the remaining 0, 4, E etc. as normal
characters.
I suspect that this is a difference between Borland and MS in
interpreting the RC file.
I have started looking at xlat to see if it will produce the data
required, which would be
2 L"Notation '{0}' has already been declared"
Is this something xlat will do just by giving it the proper arguments?
Mike
