On Thursday 26 August 2010 5:05:54 am Sean Patrick Floyd wrote: > OK, thanks for your help. > > a) Is there another way for me to generate a wsdl from java code in a build > process (without using jaxws annotations).
Probably not. You MAY be able to add a cxf.xml into src/main/resources or something that adds a simple:server or similar that would define a bunch more attributes (serviceName, endpointNmae, etc...), but I'm not really sure if that would work at all either. The JAX-WS stuff is much better tested for this case. > b) Or could you describe what's wrong in the generated wsdl and I'll write > a post-processor script that repairs it. Well, the names for the stuff in the <service> element and it's children are wrong. Well, they are correct, but the rest of the wsdl is wrong. :-) If you see the namespaces that the prefixes map to, they map to the targetNamespace you pass in. However, the targetNamespace on the wsdl:definitions is the one generated from the package. You may be able to just change the targetNamespace on the wsdl:definitions to the desired one and probably update any namespace prefix decls that refer to the generated namespace. That MAY work. Dan > > Sean > > -----Original Message----- > From: Daniel Kulp [mailto:[email protected]] > Sent: Mittwoch, 25. August 2010 17:03 > To: [email protected] > Cc: Sean Patrick Floyd > Subject: Re: Problems with maven generated services and clients > > Looks like there is a bug with: > > <targetNamespace>http://www.mycompany.com/services/test</targetNamespace> > > and the simple frontend. It's not completely honoring that so various > things > are in the wrong namespace and the generated wsdl is not correct. I would > suggest not using that at this point (and log a bug and attach your > project). > > > > <schemaTargetNamespace>http://www.mycompany.com/data/test</schemaTargetName > s pace> > <schemaTargetNamespacePrefix>td</schemaTargetNamespacePrefix> > <targetNamespacePrefix>ts</targetNamespacePrefix> > > Not sure where you got these options. They don't exist and thus would > have > > no affect. > > Dan -- Daniel Kulp [email protected] http://dankulp.com/blog
