Re: C++ interface problems

• To: Gregory Orris <orris@xxxxxxxxxxxxxxxxxx>
• Subject: Re: C++ interface problems
• From: Russ Rew <russ@xxxxxxxxxxxxxxxx>
• Date: Thu, 01 Jul 1999 13:42:13 -0600

> Date: Thu, 1 Jul 1999 12:58:57 -0400
> From: Gregory Orris <orris@xxxxxxxxxxxxxxxxxx>
> To: netcdfgroup@xxxxxxxxxxxxxxxx
> Subject: C++ interface problems

In the above message, Gregory Orris <orris@xxxxxxxxxxxxxxxxxx> wrote:

> Currently I'm trying to use the libnetcdf_c++.a interface in a fairly 
> simple program on an SGI OCTANE running IRIX 6.5.3.  While everything
> has been compiled and the tests check out fine, a bug has cropped up that
> is inexplicable:  The following lines of code 
> 
> cerr << "Is it valid?  " << output_file.is_valid() << endl;
> cerr << "What is the id?  " << output_file.id() << endl;
> cerr << "Is Latitude valid? " << output_file.get_var("Latitude")->is_valid() 
> << endl;
> 
> elicits the following output
> 
> Is it valid?  1
> What is the id?  3
> ncvarid: ncid 3: Not a netCDF id

I'm unable to duplicate the problem.  On an IRIX 6.5 system using 

 CC=/bin/c89
 CFLAGS=-g
 CXX=/bin/CC

using the MIPSpro compilers version 7.2.1.3m.

I compile and run the appended C++ program (stored in a file named
"bug1.cc") from the netcdf-3.4/src/cxx/ directory after running "make
test", so it reads the "example.nc" file created in testing.

 $ CC -o bug1 -I../libsrc -I.  -DNDEBUG bug1.cc libnetcdf_c++.a -L../libsrc 
-lnetcdf
 $ ./bug1
 Is it valid?  1
 What is the id?  3
 Is lat valid? 1

The output shown above is what is expected.  Can you send more details
to "support@xxxxxxxxxxxxxxxx" so we can reproduce the problem?
Thanks!

--Russ

#include "netcdf.hh"
int
main()
{
  NcFile nfile("example.nc", NcFile::ReadOnly);
  cerr << "Is it valid?  " << nfile.is_valid() << endl;
  cerr << "What is the id?  " << nfile.id() << endl;
  cerr << "Is lat valid? " << nfile.get_var("lat")->is_valid() << endl;
}

• References:
  • C++ interface problems
    • From: Gregory Orris