[ncl-talk] Incompatibility between addfile and isfilepresent

[Kyle Griffin]

Hi Mary,

Nope, you're the first - figured you may have had other busy/important
things along the way. I figured this was the intended use. If getting the
executable is easy, that would be great - I have some real-time code that
reads files as they are delivered and built in small pieces, so it would
stand to benefit a good bit from using both of these new implementations.

The machine in questions is RHEL5 with gcc 4.1.2.

Thanks again!


Kyle

----------------------------------------
Kyle S. Griffin
Department of Atmospheric and Oceanic Sciences
University of Wisconsin - Madison
Room 1421
1225 W Dayton St, Madison, WI 53706
Email: ksgriffin2 at wisc.edu

On Thu, Sep 18, 2014 at 1:37 PM, Mary Haley <haley at ucar.edu> wrote:

> Hi Kyle,
>
> My apologies, it appears that nobody responded to your post, unless they
> did so offline.
>
> I'm scratching my head over this bug, because I'm sure I've been using
> "isfilepresent" in the manner that you described, but yet it's not working
> for me at the moment.
>
> I consider this a serious bug and will file a report on it shortly.  If we
> get it fixed, do you want an updated "ncl" binary with the fix?  I would
> only need to give you a new "ncl" executable.
>
> --Mary
>
> On Tue, Sep 9, 2014 at 4:55 PM, Kyle Griffin <ksgriffin2 at wisc.edu> wrote:
>
>> Just setting up 6.2.1 today, and finding that files that CAN be opened
>> with addfile does not register with isfilepresent. While I could use
>> filexists to duplicate the old functionality of the command, it would be
>> nice to be able to take advantage of the "is this an openable file" check
>> before feeding it to addfile later in my script.
>>
>> Specifically, the files that are causing this issue are ones that are
>> GRIB2 files but do not contain a file extension. While I can give it the
>> appropriate file extension to make addfile work (it checks for
>> 'file_name.grb2' as well as simply 'file_name'), this feature does not seem
>> to exist with isfilepresent and the function returns false for all calls
>> with or without file extensions.
>>
>> I feel this isn't the desired implementation of this change to the
>> function, and I think it might be an easy fix. Any insights? I've included
>> some command-line interaction with the file below (specific file doesn't
>> matter, as this issue should be true for any GRIB or GRIB2 file that is on
>> the file system but has no extension).
>>
>> Let me know if I can/should provide further evidence, and thanks!
>>
>>
>> Kyle
>>
>>
>>
>> ncl 0>
>> print(isfilepresent("/cas2/unidata/GRIB/gfs0p5deg/GFS_140909_18_000.grb"))
>>
>> fatal:NetCDF: Unknown file format
>>
>> fatal:Could not open (/cas2/unidata/GRIB/gfs0p5deg/GFS_140909_18_000)
>>
>> (0) False
>>
>> ncl 1>
>> print(isfilepresent("/cas2/unidata/GRIB/gfs0p5deg/GFS_140909_18_000.grb2"))
>>
>> fatal:NetCDF: Unknown file format
>>
>> fatal:Could not open (/cas2/unidata/GRIB/gfs0p5deg/GFS_140909_18_000)
>>
>> (0) False
>>
>> ncl 2>
>> a=addfile("/cas2/unidata/GRIB/gfs0p5deg/GFS_140909_18_000.grb","r")
>>
>> ncl 3> printVarSummary(a)
>>
>>
>> Variable: a
>>
>> Type: file
>>
>> File path: /cas2/unidata/GRIB/gfs0p5deg/GFS_140909_18_000
>>
>> Number of global attributes: 0
>>
>> Number of dimensions: 15
>>
>> Number of variables: 108
>>
>> ncl 4>
>> print(isfilepresent("/cas2/unidata/GRIB/gfs0p5deg/GFS_140909_18_000"))
>>
>> fatal:NetCDF: Unknown file format
>>
>> fatal:Could not open (/cas2/unidata/GRIB/gfs0p5deg/GFS_140909_18_000)
>>
>> (0) False
>> ----------------------------------------
>> Kyle S. Griffin
>> Department of Atmospheric and Oceanic Sciences
>> University of Wisconsin - Madison
>> Room 1421
>> 1225 W Dayton St, Madison, WI 53706
>> Email: ksgriffin2 at wisc.edu
>>
>> _______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20140918/5f2a759b/attachment.html