On Wednesday, October 19, 2011 5:19:17 AM acollier wrote: > We are using aegis data binding to produce SOAP and wsdl content in front a > web server. Integration with spring works great, and the services are > consumable by both Java and PHP clients. But if I validate the SOAP response > with a tool (such as SOAP UI), an error occurs on complex type binding > involving an inheritance pattern. > > Complex inheritance scheme: > > AbstractObjectA -implements-> InterfaceObjectA > ObjectAImpl -extends-> AbstractObjectA > AbstractObjectB -implements-> InterfaceObjectB > AbstractObjectB -extends-> AbstractObjectA > ObjectBImpl -extends-> AbstractObjectB > > Spring and aegis configuration: > > > here is the wsdl:types part of resulting wsdl : > > > we can see here that inheritance between interfaces and abstract objects are > not described.
Well, I guess the question is "what *should* it look like in the schema?" I guess, most likely, the "InterfaceObjectA/B" types should just not be there. Not 100% sure though. Maybe some sort of complex element substitution things, but that would really get messy. > The inheritance tree isn't complete and similar to what is > coded in java. If a service return type is described with an interface and > the real content is an implementation object, validation of soap response > will fail. Java clients (cxf aegis client) seem to be smart enough to > unmarshall that kind of inheritance tree, PHP clients need to declare a > data type mapping to complete the inheritance tree, but .NET client are > blocking on validation of the soap response as they fail to match interface > and implementation types. > > Does the inheritance tree is too complex for the aegis data binding process? > Is there option we can tune to make it work? > Is it a nominal behaviour? I wouldn't say "normal behavior", but you likely have gone past the regular complexity level so more into "uncharted waters". Dan > > Thanks, > > Alex > > > -- > View this message in context: > http://cxf.547215.n5.nabble.com/aegis-databing-SOAP-response-does-not-respe > ct-validation-standard-tp4917484p4917484.html Sent from the cxf-user mailing > list archive at Nabble.com. -- Daniel Kulp [email protected] http://dankulp.com/blog Talend - http://www.talend.com
