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

Re: Problem handling NULL bytes in TDS

Weve decided that we should indeed truncate any Strings if they contain a null
byte. I will put that
fix into both the 2.22 branch and the trunk. Kevin, do you need this fix right
away?

Kevin O'Brien wrote:

Hi John -

It is indeed a netcdf file, albeit with messed up attributes.  You can
download the file at



http://nomads.gfdl.noaa.gov/dods-data/gfdl_cm2_1/CM2.1U-D4_1860-2000-AllForc_H1/pp/ice_tripolar/ts/monthly/vsi_O1.186101-200012.nc



I could send you the header, but I'm not sure how the NULL bytes would
get translated along the way...But let me know if you want that or if
it's just easier to download the file.  I think it's about 500Mbytes...

kev

John Caron wrote:

Hi Kevin:

Theres some question about whether this is the right thing to do or
not. Is this a netcdf file? If so, can you send it to me?

Kevin O'Brien wrote:

Hi John -

I ran across what looks to potentially be a bug in the way TDS
handles NULL bytes within attributes.  If you plug this URL into the
netCDF Java tools interface:
     http://data1.gfdl.noaa.gov:8380/thredds/dodsC/test_vsi_monthly

And hit the ncdump tab,  and then scroll down to the VSI variable,
you'll see the null bytes in the
   cell_methods
and
   coordinates
attributes.   I was talking to Roland about this, and he seems to
think that the attribute value should end w/ the NULL byte.
Also, if you then hit the "NcML" tab, you'll get an error complaining
about the NULL byte, though in a different attribute (grid_version in
this case).

If there's any more information I can provide, please let me know..

thanks-
Kevin