Re: confusion about H5Gget_objname_by_idx

Hi Ed,
    Yup, you are correct - I'll report it as a doc bug.

        Quincey

> Seems like an error in the docs. The ref manual sez: (note last parameter)
> 
> 
> Name: H5Gget_objname_by_idx
> Signature:
>     ssize_t H5Gget_objname_by_idx(hid_t loc_id, hsize_t idx, char
> *name, size_t* size ) 
> 
> 
> But, H5Gpublic.h contains the following:
> 
> H5_DLL ssize_t H5Gget_objname_by_idx(hid_t loc_id, hsize_t idx, char* name, 
> size_t size);
> 
> 
> The last parameter is not a pointer after all!
> 
> Ed
> 
>From owner-netcdf-hdf@xxxxxxxxxxxxxxxx 01 2003 Nov -0700 09:44:49 
Message-ID: <wrxllr0f37i.fsf@xxxxxxxxxxxxxxxxxxxxxxx>
Date: 01 Nov 2003 09:44:49 -0700
From: Ed Hartnett <ed@xxxxxxxxxxxxxxxx>
To: netcdf-hdf@xxxxxxxxxxxxxxxx
Subject: extensible datasets in every dimension?
Received: (from majordo@localhost)
        by unidata.ucar.edu (UCAR/Unidata) id hA1GipIJ017962
        for netcdf-hdf-out; Sat, 1 Nov 2003 09:44:51 -0700 (MST)
Received: from rodney.unidata.ucar.edu (rodney.unidata.ucar.edu 
[128.117.140.88])
        by unidata.ucar.edu (UCAR/Unidata) with ESMTP id hA1GioOb017957
        for <netcdf-hdf@xxxxxxxxxxxxxxxx>; Sat, 1 Nov 2003 09:44:50 -0700 (MST)
Organization: UCAR/Unidata
Keywords: 200311011644.hA1GioOb017957
Lines: 32
User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: owner-netcdf-hdf@xxxxxxxxxxxxxxxx
Precedence: bulk

Greeting to HDF HQ!

Some questions:

1 - If any dimension of a dataset is to be extensible, the dataset
must be chunked, correct?

2 - If I have a dataset with one expendables dimension, and several
fixed dimensions, I can create it using a space like this:

      if ((spaceid = H5Screate_simple(var->ndims, dimsize, maxdimsize)) < 0)
         BAIL(NC_EHDFERR);

Where dimsize is an array of initial dimension sizes, and maxdimsize
is an array of max dimension sizes, or -1 for an unlimited dimension.

The problem I have is this. If I try and create a space with one of
the dimsizes as zero, and the corresponding maxdimsize as -1, HDF5
hands when I try to close the defined dataset (this seems like a HDF
bug, BTW).

So I can do it if I start with a dimsize of 1 instead of 0, but this
is not the netcdf way. After defining such a dataset, when I get the
number of records in it, using H5Sget_simple_extent_dims, I get a size
of one along that dimension, even though I haven't written any data to
it yet, because I have to define it as having length of at least one.

Any comment or help would be appreciated...

Thanks!

Ed