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

[THREDDS #WFC-957660]: Similar functionality in THREDDS to OPeNDAP FreeForm + constraint expressions?



Hi Peter,

The plan has been to start using the THREDDS DQC services to replace the 
OPeNDAP File Server functionality. Which would mean the queries are not done as 
OPeNDAP constraints but as DQC queries instead.

We haven't had the resources to work on the DQC much in the last year or two. 
So, we are still at the stage of trying to figure out how this switch should 
work and what changes might need to be made to the DQC spec and our DQC code 
before a switch (or migration) could be made. The DQC spec is a bit more 
advanced than our DQC service code which is more at a prototype level. As we 
figure out how to proceed, both the spec and the code will need some 
development. Here is the DQC page:

http://www.unidata.ucar.edu/projects/THREDDS/tech/dqc/DqcStatus.html

Let me know if you have any DQC questions. Also, if you have any thoughts on 
this matter, we would be interested in hearing them.


Other possibilities:

1) As far as NcML: The XML Schema for NcML does support variable length arrays 
(see dimension attribute "isVariableLength") which map to OPeNDAP sequences. 
However, I'm not sure how an NcML dataset served by the THREDDS Data Server 
(TDS) via OPeNDAP would look. In other words, I'm not sure sequences are fully 
supported in all our code yet or if the query side of things is supported. Are 
you running a THREDDS Data Server (TDS)? That is the only server that could 
serve an NcML file as a dataset.

If it did work this would be another route to an OPeNDAP File Server besides 
FreeForm.

2) You could build a THREDDS InvCatalog 
(http://www.unidata.ucar.edu/projects/THREDDS/tech/catalog/InvCatalogSpec.html) 
for your collection and include the appropriate metadata that you could search 
on. This is one possible backend behind a DQC document/service interface. 

3) Another possibility is aggregation of your collection into a single dataset. 
The TDS also supports dataset aggregation (which uses NcML to describe the 
aggregation). However, the TDS only supports HDF5 (not HDF4) and only a subset 
of HDF5 files (also not sure if the HDF5 support is at alpha or beta support). 
So, not sure it would work for your particular HDF files.


Anyway, hope this was helpful. Sorry it is fairly rambling. Let me know if you 
have other questions.

Ethan




> Hi support,
> 
> I was going through the OPeNDAP FreeForm ND Data Handler User's Guide
> at:
> 
> http://www.opendap.org/server/dff-html/dff.html
> 
> and specifically the File Servers section (Section 7) at:
> 
> http://www.opendap.org/server/dff-html/dff_42.html
> 
> when I read the part about using THREDDS rather than an OPeNDAP
> server with FreeForm file as the index.  What I can't understand from
> the THREDDS documentation is, what parts of THREDDS may be used to
> supply the same functionality as you get with an OPeNDAP server and a
> FreeForm index on which I can use constraint expressions to select
> datasets of interest?
> 
> For example, I have a directory full of HDF satellite data files,
> each one from a specific date and time and containing a single 2D
> grid of values.  Each data file represents the same geophysical area
> on the earth.  Say that for each one, I've extracted the date, time,
> ground station, percent of valid pixels in the grid, and various
> other key pieces of metadata that I want to use as criteria in an
> OPeNDAP constraint expression.  I also have the OPeNDAP URL that I'm
> going to use in my client program after I get the search results.  I
> can't see from the THREDDS documentation how I can perform that type
> of search.  Is it possible?
> 
> I thought of maybe using NcML to create an index dataset that
> contained OPeNDAP "sequence" data that I could run a constraint
> expression on, but I can't see that NcML supports sequences
> specifically, unless your OPeNDAP implementation translated groups
> into sequences in special cases?  Or is there some way that I can
> make use of the THREDDS Dataset Query Capability (DQC) Services?
> 
> Peter
> 
> --
> Peter Hollemans, SP Systems for NOAA/NESDIS


Ticket Details
===================
Ticket ID: WFC-957660
Department: Support THREDDS
Priority: Normal
Status: Open


NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.