Hi Dan, Yes, turn on validate reject it. Thanks
Freeman －－－－－－－－－－－－－ Freeman Fang FuseSource Email:[email protected] Web: fusesource.com Twitter: freemanfang Blog: http://freemanfang.blogspot.com http://blog.sina.com.cn/u/1473905042 weibo: http://weibo.com/u/1473905042 On 2012-9-3, at 下午8:39, [email protected] wrote: > Freeman > > It's an invalid WSDL and should be rejected by the code generator. If you run > with validation turned on, does it reject it? > > With WSDL1, operations without an output cannot throw faults. > > Daniel Kulp > http://dankulp.com/blog > > > On Sep 3, 2012, at 4:43 AM, Freeman Fang <[email protected]> wrote: > >> Hi, >> >> Currently when we use wsdl first way, if an operation in the wsdl have input >> and fault but no output, then it would be considered as a oneway operation, >> we can see a auto-generated oneway operation can throw exception. >> Is this behavior correct? I doubt it but I checked the JAXWS spec but can't >> find clear definition about this part. >> >> When use code first way, if a java method has Oneway annotation, then it's >> fault would be ignore, you can see it from ReflectionServiceFactoryBean >> if (hasOut) { >> // Faults are only valid if not a one-way operation >> initializeFaults(intf, op, method); >> } >> >> So the way we process oneway is different between wsdl-first and code-first. >> In wsdl-first, we keep the fault but in code-first we ignore the fault. This >> mismatch could cause problem like CXF-4493, I think we should unify oneway >> behavior between wsdl-first and code-first. >> >> Any thoughts? >> >> Thanks >> Freeman >> －－－－－－－－－－－－－ >> Freeman Fang >> >> FuseSource >> Email:[email protected] >> Web: fusesource.com >> Twitter: freemanfang >> Blog: http://freemanfang.blogspot.com >> http://blog.sina.com.cn/u/1473905042 >> weibo: http://weibo.com/u/1473905042 >> >>
