Re: netcdf-4.0-beta1 release now available on netCDF web site...

Jeff Whitaker <jswhit@xxxxxxxxxxx> writes:

>
> Ed:  Thanks!  beta1 works fine and passes all the tests with my python
> module.  However, ncdump appears to be broken - every netcdf-4 file I
> try results in something like this:
>
> ncdump: test.nc: HDF error
>
> It seems to work fine with netcdf-3 files however.
>

Howdy Jeff!

In this release, ncdump and ncgen can handle netCDF-4 files as long as
they conform to the classic model.

That is, as long as you don't use user-defined types and the new
atomic types, it should work fine. (Ncdump does handle groups
correctly, but not yet ncgen.)

This is tested with the ncdump tests, when the script tst_netcdf4.sh
is run in the ncdump directory. This script causes ncgen to create
some netCDF-4/HDF5 files from the same CDL scripts that are used to
test the netCDF-3 ncgen/ncdump stuff. Then the script feeds the
resulting data files back to ncdump to make sure that the same CDL
output results.

For example:

bash-3.00$ ../ncgen/ncgen -v3 -b -o c0.nc ../ncgen/c0.cdl
bash-3.00$ ./ncdump c0.nc 
netcdf c0 {
dimensions:
        Dr = UNLIMITED ; // (2 currently)
        D1 = 1 ;
        D2 = 2 ;
        D3 = 3 ;
        dim-name-dashes = 4 ;
        dim.name.dots = 5 ;
variables:
        char c ;
                c:att-name-dashes = 4 ;
                c:att.name.dots = 5 ;
<etc.>

Do your files conform to the classic model?

Dealing with the new netcdf-4 features in ncdump is one of our
priorities for the next beta release, but partial functionality is
there now.

Thanks!

Ed
-- 
Ed Hartnett  -- ed@xxxxxxxxxxxxxxxx

==============================================================================
To unsubscribe netcdfgroup, visit:
http://www.unidata.ucar.edu/mailing-list-delete-form.html
==============================================================================