Hi,
On 29 Oct 2003 19:36:57 +0800, Gerrit Voss <[EMAIL PROTECTED]>
wrote:
Hi.
On Wed, 2003-10-29 at 19:01, Haykel Ben Djemia wrote:
Hi,
in OpenSG/Source/Experimental/NewAction/ActorCodeGen/OSGActorWriter.cpp
string::compare(int, int, char*) is used, which is OK for new libstdc++
releases under linux. But the libstdc++-2.96 delivered with RedHat 7.3
has the following definition: string::compare(char*, int, int). For
support reasons for our E&S cluster we have to use RedHat 7.3. For the
moment I compiled OpenSG by changing every strLine.compare(0, n, "text")
with strLine.compare("text"). Could you make these changes in the
OpenSG-Source and take care of it in the future or is this problematic
for some reason?
no reason I can see without having checked that the first parameter is
always 0 ;-) if this is the case Carsten will handle the details ;-)).
alright, the first arg is always 0, but I don't like the proposed solution.
Consider this:
if(strLine.compare(0,5, "@@Bar") == 0) // this means the line starts
exactly with @@Bar
vs.
if(strLine.compare("@@Bar") >= 0) // this means the line starts with
anything lexicographically ordered after @@Bar, e.g. @@Foo
I don't like testing with == in the second case either, as this would not
work on lines that have trailing whitespace for example.
If not it will be a little bit trickier as I don't like to sprinkle the
code with a lot of ifdefs ;-)
Well, ActorWriter already has a workaround member function for the missing
strcasecmp on win32, I'd add another one for the string::compare problem,
but I don't know which defines to test.
Carsten
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users
