I've seen traffic on null-termination of strings + Fortran + netCDF, and
have read the section in the manual regarding this issue. BUT, I thought I
understood this to mean that attribute strings would not be terminated by a
null.
The files I am reading are written by a python-netCDF processor, and the
strings I get from attributes using the Fortran interface (netCDF version
3.4) are null terminated and the length I get back from nf_inq_attlen
includes the null. If this is the standard, expected behavior, I can
simply look for and remove null characters any time I read an attribute
string. I would rather, however, that something a bit lower down in the
library chain deal with this for me. Am I missing something?
Jim
*-------------------*-------------------------------*
| Jim Secan | Northwest Research Assoc, Inc |
| (jim@xxxxxxxx) | 9040 S Rita Road, Suite 2214 |
| (520) 663-3570 | Tucson, Arizona 85747 |
| | http://www.nwra.com/ |
*-------------------*-------------------------------*
