[seek-dev] Kepler/Garp testcase needed

[Matt Jones]

OK, so I spoke with Jing about the "content" element in the EcoGrid 
resultset.xsd schema, and we agreed it should be removed.  It was only 
inserted because of a bug in the Axis stub generation tools, and we 
agreed we didn't want platform/language bugs to drive the development of 
the web service interfaces.  So, Jing will remove it fromthe xsd file 
and make sure all of the test files validate.  Then he will modify the 
stub code manually to work around the bug.  Jing and Bing will modify 
the metacat and srb wrappers to return the resultset according to this 
modified schema.  Bing -- let us know if this will be problematic for 
you.  Rod can program the DiGIR wrapper to the modified schema that 
doesn't use the "content" element.

See IRC log below for our conversation,

Matt

IRC Log for #seek, 2004-01-20, 14:52:00
---------------------------------------
<jing> hi, matt.
<matt> hey
<jing> yep, I added content type
<jing> The reseason is simple,
<jing> previsous version, AnyRecordType is base on anyType.
<jing> if in this way, the attributes, for example, identifier,
<jing> would not be shown in java file.
<jing> This is bug in Axis
<matt> just in the axis stub generation code?
<jing> yes.
<matt> then i think we should manually make the changes
<jing> Yes, definitely we can do it.
<matt> we shouldn't have to make the schema more complex just to 
accomodate one
soap package
<matt> what if there'sa bug in the perl stub generators?
<matt> or python?
<matt> or .NET
<matt> if we modify schemas for each, we lose our language/platform 
independence<matt> the stub generator is just a convenience.
<jing> you are  right.
<matt> what do the metacat and srb wrappers for ecogrid produce now?
<jing> the current resultset.xsd
<matt> is it anything more than trivial to modify them?
<matt> ie, can it be doen for both systems in < 1 hour
<jing> no, I don't think so.
<jing> less thank 1 hour?
<matt> how long would it take?
<jing> I am not sure.
<matt> hours, days, months :)
<jing> hours
<jing> you know, we need testing.
<matt> yes, i know.  its on rods list
<matt> well, rod is developing the digir client now
<matt> what should i recommend to him: with or without the "content" element
<matt> ?
<jing> For metacat it is fine without content,
<jing> I can fix it quickly
<jing> but I think we need to talk with bing, so see how about srb
<matt> ok. can you do that, and resolve this asap so rod can make progress?
<jing> sure.
<matt> the xsd file will need to be updated, and make sure all of the 
test files validate properly
<matt> *all*
<matt> :)
<jing> yep
<matt> ok, thanks.  i'm going to send this irc log to seek-dev so they 
know...
-----------------------------------


Matt Jones wrote:
> Rod,
> 
> I put a test digir ecogrid query and resultset document in 
> seek/projects/ecogrid/tests/testfiles.  You should be able to run a 
> Digir query against:
>   http://speciesanalyst.net/digir/DiGIR.php?resource=MammalsDwC2
> and get back results.  I'm not sure which mammal species in there have 
> lat/lon values -- you'll need to find one that does.  Dave can probably 
> help with that, or tell us an alternate resource that does have lat/lon 
> values.
> 
> Let me know if they are what you had in mind, and see my CVS log entries 
> for some issues that I came across while doing this.  In particular, it 
> appears a change to the resultset.xsd was made (myabe by Jing?) to add a 
> "content" element that I think is superfluous, but we'll see by 
> discussing it with Jing.  So my test resultset probably won't validate 
> against the schema until this is resolved.
> 
> Matt
> 
> Rod Spears wrote:
> 
>> Can somebody put together an xml document containing the Ecogrid query 
>> that represents the "digir" query that will be used inside of Kepler 
>> for the Beam demo?
>>
>> I need a realistic tesctase to make sure I have the transform correct, 
>> also if you could attach the what a result set would look like that 
>> would be a big help also.
>>
>> Thanks,
>> Rod
>>
>> -- 
>> Rod Spears
>> Biodiversity Research Center
>> University of Kansas
>> 1345 Jayhawk Boulevard
>> Lawrence, KS 66045, USA
>> Tel: 785 864-4082, Fax: 785 864-5335
>>
> 

-- 
-------------------------------------------------------------------
Matt Jones                                     jones at nceas.ucsb.edu
http://www.nceas.ucsb.edu/    Fax: 425-920-2439    Ph: 907-789-0496
National Center for Ecological Analysis and Synthesis (NCEAS)
University of California Santa Barbara
Interested in ecological informatics? http://www.ecoinformatics.org
-------------------------------------------------------------------