Looks like someone already noticed this and there's some movement on it. http://nagoya.apache.org/bugzilla/show_bug.cgi?id=20109
-----Original Message----- From: Cory Wilkerson Sent: Monday, September 08, 2003 9:41 AM To: '[EMAIL PROTECTED]' Subject: a consensus on xsd/soapenc int? All, Assume xsd = http://www.w3.org/2001/XMLSchema Assume soapenc = http://schemas.xmlsoap.org/soap/encoding I don't think we ever came to a conclusion on whether or not Axis is behaving improperly when it qualifies ints in the "soapenc" namespace when generating stubs from WSDL (rpc/encoded) that indicate the int belongs in the xsd namespace (but has a minOccurs of 0), e.g. (paraphrased): <xsd:schema targetNamespace="http://www.foo.com"; xmlns:xsd="http://www.w3.org/2001/XMLSchema";> <xsd:complexType name="foo"> <xsd:sequence> <xsd:element name="bar" type="xsd:int" minOccurs="0"/> </xsd:sequence> </xsd:complexType> </xsd:schema> WSDL2Java generates the following, rather unhealthy code IMHO: field = new org.apache.axis.description.ElementDesc(); field.setFieldName("bar"); field.setXmlName(new javax.xml.namespace.QName("", "bar")); field.setXmlType(new javax.xml.namespace.QName("http://schemas.xmlsoap.org/soap/encoding/";, "int")); field.setMinOccursIs0(true); I managed to find where this assumption is made in the WSDL2Java code -- my question is -- is that assumption correct or incorrect? I've got all my money on it being incorrect but would like a little validation. Thanks, Cory
