[MISC] [MOBY-l] Help on writing .NET client for Moby services

Mark Wilkinson markw at illuminae.com
Mon Jun 21 15:47:18 UTC 2004


On Sun, 2004-06-20 at 20:31, Gerald Sylvester wrote:

> Greetings from Gerard.

Hi Gerard!


> (1) How do I access the Moby Central from my .NET
> client and generate the WSDL file, does the services
> registered there use the .asmx file extension?

Could you clarify which WSDL file you require - a WSDL representing MOBY
Central, or a WSDL representing the service you wish to invoke?

Services don't use any predictable file extension, and may or may not
use a file extension at all.

 
> (2) I need to get the WSDL file inorder to write the
> client code for that particular webservice.

Which parts of the WSDL are important to you?  If it is just the
endpoints, then you can get sufficiently complex WSDL from the MOBY
Central retrieveService call.  If you need the XSD describing service
inputs/outputs, then... you are out of luck... unless you want to write
the code that generates it ;-)  The *actual* input/output structure of a
service is relatively unpredictable, but it is predictable enough to
build a client/service around.  Since we use an ontology of data-types,
and the ontology dictates the XML structure, we can rely on the XML
being *at least* as complex as that described by the ontology node that
is registered in MOBY Central as the input or output for that service. 
However, it may be even more complex (hence it is very difficult to
write a valid WSDL document describing the service...).  Regardless, if
you use an XML parser to reach into the XML you receive, based on the
ontology node you are expecting, and pull out the various nodes that
should be there, you will find that they are there.  If you are a
particularly clever client, you can try to make sense of the entire
message, but you do not need to do so.  

Having said that, the fact that your client cannot make sense of a
message does not stop you from passing this message on to another
service provider who claims to be able to do something with it!  Hence,
pipelining is not thwarted by the clients inability to make sense of
things...

M


-- 
Mark Wilkinson (mwilkinson at mrl.ubc.ca)
University of British Columbia iCAPTURE Centre



More information about the moby-l mailing list