Hi,
I've been having some issues copying datasets; in some cases the new
dataset will be created but will be empty (no storage allocated).
I've finally tracked it down to an open identifier; if you create a
dataset, write to it and then call H5Ocopy, a new dataset will be
created but it will be empty.
This is a problem for my application (h5py) because I don't directly
control the lifecycle of the identifiers; my users do. In any case
this behavior isn't documented.
I've attached a C file which demonstrates the problem. Any ideas?
Thanks,
Andrew Collette
test.c (2.12 KB)
Andrew,
This is a bug. Thank you for reporting! I will add it to our bugs for the HDF5 1.8.9 release. We will let you know when we have a fix.
Elena
ยทยทยท
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Elena Pourmal The HDF Group http://hdfgroup.org
1800 So. Oak St., Suite 203, Champaign IL 61820
217.531.6112
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
On Nov 28, 2011, at 11:17 PM, Andrew Collette wrote:
Hi,
I've been having some issues copying datasets; in some cases the new
dataset will be created but will be empty (no storage allocated).
I've finally tracked it down to an open identifier; if you create a
dataset, write to it and then call H5Ocopy, a new dataset will be
created but it will be empty.
This is a problem for my application (h5py) because I don't directly
control the lifecycle of the identifiers; my users do. In any case
this behavior isn't documented.
I've attached a C file which demonstrates the problem. Any ideas?
Thanks,
Andrew Collette
<test.c>_______________________________________________
Hdf-forum is for HDF software users discussion.
Hdf-forum@hdfgroup.org
http://mail.hdfgroup.org/mailman/listinfo/hdf-forum_hdfgroup.org