[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 961022: extracting data from improperly closed netCDF

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: Re: 961022: extracting data from improperly closed netCDF
  • Date: Tue, 22 Oct 1996 17:13:59 -0600
Becky:

Most likely all you need to do is update the 'numrecs' field as Russ outlines
in his first suggestion. Then you should be able to read the file normally.
You shouldn't need to parse things further than this, as in Russ's second
suggesstion. Just change the second 4 byte bigendian word to a number greater
than or equal to the number of records actually in the file.
Then read the file as usual. You will get an error when you try to read beyond
the end of file.

If you have a version of the netcdf library which was compiled using
the '-g' flag for debugging, you can read off the header information
and examine it in the debugger. Stop at the last executable line of NC_open().
Then "print *handle".
 You can use 'handle->begin_rec', 'handle->recsize' and the size of the
 file (from 'ls -l') to compute the number of records:
numrecs = (file_size - begin_rec)/recsize.

Some debuggers (SGI 'cvd', for example) will allow you to change values there
in the debugger. You could set handle->numrecs and proceed...

The symbols above apply to netcdf-2.

The last record to be written is probably garbage.

-glenn