This is not the place where serialization happens. The code you've pointed
out is used when generating the WSDL 2.0 doc for your service (when ?wsdl2
is invoked).
Thanks,
Keith.
On Tue, Jun 17, 2008 at 5:22 PM, <[EMAIL PROTECTED]> wrote:
>
> Hello,
>
>
> I uses Axis2-1.4's wsdl2java to generate a client to a web service that
> does *not* accept the "mustUnderstand" attribute in headers. Obviously,
> sending it headers with mustUnderstand attribute set to something (true or
> fals or 0 or 1) fails. So, I want to make sure that the generated client
> doesn't add the dreaded attribute. I've checked this ML and there are
> several people who contributed information on how to achieve that. But none
> of the proposed solutions seem to work for me. Here's an excerpt of my code,
> with all the solutions I've tested :
>
> public static void main(String[] inArgs) throws
> AxisFault,AlineFault,RemoteException,Exception {
>
> AlineBackendServiceStub ws = new
> AlineBackendServiceStub("http://...";);
> // address hidden for privacy; that's the stub, coming right from wsdl2java
>
> Map m = ws._getServiceClient().getOptions().getProperties();
>
> m.remove(AddressingConstants.ADD_MUST_UNDERSTAND_TO_ADDRESSING_HEADERS);
> ws._getServiceClient().getOptions().setProperties(m);
>
>
> //ws._getServiceClient().getOptions().setProperty(AddressingConstants.ADD_MUST_UNDERSTAND_TO_ADDRESSING_HEADERS,
> Boolean.FALSE);
>
> ws._getServiceClient().getServiceContext().getConfigurationContext().setProperty(AddressingConstants.ADD_MUST_UNDERSTAND_TO_ADDRESSING_HEADERS,
> Boolean.FALSE);
>
> ws._getServiceClient().getServiceContext().getConfigurationContext().createMessageContext().setProperty(AddressingConstants.ADD_MUST_UNDERSTAND_TO_ADDRESSING_HEADERS,
> Boolean.FALSE);
>
> ...
>
> }
>
> None of these work : the mustUnderstand attribute kept written in my header
> (with value "0").
>
> I've also tried to interfere with message context in the generated stub,
> like this :
>
> // create a message context
> _messageContext = new
> org.apache.axis2.context.MessageContext();
>
>
> _messageContext.setProperty(AddressingConstants.ADD_MUST_UNDERSTAND_TO_ADDRESSING_HEADERS,
> Boolean.FALSE); // Added by me !!!
>
>
> Then, I 'depserately) looked at the source code of Axis, and the only thing
> I've seen is this (in WSDLSerializationUtil.java):
>
>
> /**
> * Adds a soap header element to a given OMElement
> * @param omFactory - An OMFactory
> * @param list - The arraylist of soapHeaderMessages
> * @param wsoap - The WSDL 2.0 SOAP namespace
> * @param element - The element that the header should be added to
> * @param nameSpaceMap - The namespaceMap
> */
> public static void addSOAPHeaderElements(OMFactory omFactory, ArrayList
> list, OMNamespace wsoap,
> OMElement element, Map
> nameSpaceMap) {
> for (int i = 0; i < list.size(); i++) {
> SOAPHeaderMessage soapHeaderMessage = (SOAPHeaderMessage)
> list.get(i);
> OMElement soapHeaderElement =
>
> omFactory.createOMElement(WSDL2Constants.ATTRIBUTE_HEADER, wsoap);
> QName qName = soapHeaderMessage.getElement();
> soapHeaderElement.addAttribute(omFactory.createOMAttribute(
> WSDL2Constants.ATTRIBUTE_ELEMENT, null,
> getPrefix(qName.getNamespaceURI(), nameSpaceMap) + ":"
> + qName.getLocalPart()));
> soapHeaderElement.addAttribute(omFactory.createOMAttribute(
> WSDL2Constants.ATTRIBUTE_MUST_UNDERSTAND, null,
>
> Boolean.toString(soapHeaderMessage.isMustUnderstand())));
> soapHeaderElement.addAttribute(omFactory.createOMAttribute(
> WSDL2Constants.ATTRIBUTE_REQUIRED, null,
> Boolean.toString(soapHeaderMessage.isRequired())));
> element.addChild(soapHeaderElement);
> }
> }
>
>
> I've no idea if this is the place where the serialization does happen, but
> if it is, then I believe the mustUnderstand attribute will always be set (to
> true or false) and never removed.
>
> I know this question has been posted several times and I hope that some of
> you will have enough patience to look at it...
>
> Thanks,
>
> Stefan
>
>
>
--
Keith Chapman
Senior Software Engineer
WSO2 Inc.
Oxygenating the Web Service Platform.
http://wso2.org/
blog: http://www.keith-chapman.org
