I am resending the code below, without the '*'
John Caron wrote:
Adityarajsingh Santokhee wrote:
Hello,
I have a few questions concerning reading GRIB files using the netcdf2.2
library:
1. are there any problems when using the grib.jar library with
jakarta tomcat
server ? Netcdf files can be correctly read and data extracted. Grib
files can
be opened only but it is failing each time i wish to extract data
(all data as
well as section of it).
most likely problem is you dont have write permission in the data
directories, since we need to write an index file.
however, on failure it should have defaulted over to writing in a
temporary directory. probably theres no home directory for the tomcat
user.
If you cant fix those 2 problems, set the directory to something that
the tomcat user has write privilege in, using:
ucar.nc2.util.FileCache.setRootDirectory()
2. Are there any easier ways to reading data for a particular
variable using its
internal name instead of the variable name. Currently, I have to get
list of
all variables and then for each variable get its attributes and
thereafter
check the attribute name for GRIB_param_number.
I would make a HashMap
HashMap gribHashMap = new HashMap();
for (int i=0; i<myVarList.size(); i++) {
Variable myVar = (Variable) myVarList.get(i);
String gribName = ncfile.findAttValueIgnoreCase(myVar,
"GRIB_param_number", null);
if (gribName != null)
gribHashMap.put( gribName, myVar);
}
Then look it up via
Variable myVar = (Variable) gribHashMap.get( gribName);
if ( myVar != null) {
do_something();
}
