Hi Ethan,
Thanks for this. I'm not an expert on the MyOcean file format -
Tony/Thomas, can you answer Ethan's question? Is the run time just
encoded in the file name or also in a global attribute?
Ethan, if there are two date/times in the filename and the first
date/time indicates the run time, can this be extracted by TDS, i.e. can
the second date/time be ignored? If so, what's the file pattern we'd
need to include in the TDS catalogs?
Cheers, Jon
-----Original Message-----
From: myocean-tools@xxxxxxxxxxxxxxxx
[mailto:myocean-tools@xxxxxxxxxxxxxxxx] On Behalf Of Ethan Davis
Sent: 26 July 2010 23:57
To: thredds@xxxxxxxxxxxxxxxx
Cc: myocean-tools@xxxxxxxxxxxxxxxx
Subject: [myocean-tools] Re: [thredds] FMRC aggregations with two dates
in the filename
Hi Jon,
The FMRC code always determines the valid time of the data by opening
the data file and looking at the time coordinates for the data. The run
time, on the other hand, needs to be given either in the file name or
elsewhere in the data, in a global attribute for instance.
How is the run time encoded in these files? Can you send us some example
datasets?
Thanks,
Ethan
On 6/22/2010 3:02 PM, Jon Blower wrote:
> Hi,
>
> We have a potential issue in MyOcean that filenames typically contain
> two dates, representing the start and end validity times of data
within
> the file. When creating an FMRC aggregation the filename is usually
> parsed to get the run time, but this parsing is hard when the file
name
> contains two dates/times, both of which vary between files, e.g.
>
"ice_drift_polstereo_wsm_envisat_north_20100601013759-20100602010520.nc"
> .
>
> Is there a workaround for this? See
> http://www.resc.reading.ac.uk/trac/myocean-tools/ticket/18.
>
> Cheers, Jon
>
> --
> Dr Jon Blower
> Technical Director, Reading e-Science Centre
> Environmental Systems Science Centre
> University of Reading
