According to the process of creation AXISService we should parse WSDL file and all imported schemas. But in case of big WebServices it takes a lot of memory(aproximately 200-300 Mb) Is there any abilities to exclude processing of schemas?
I parse WSDL by ibm implementation of javax.wsdl.*
Sample of code:
public static Definition read(FileObject wsdlFile)
throws Exception
{
try {
WSDLReader reader = WSDLFactory.newInstance().newWSDLReader();
reader.setFeature("javax.wsdl.verbose", false);
reader.setFeature("javax.wsdl.importDocuments", true);
return reader.readWSDL(new UnifiedURIResolver(wsdlFile));
} finally {
VfsUtils.close(wsdlFile);
}
}
solution is process xsd into UnifiedURIResolver class (implemenation of URIResolver) and if we get import of xsd we should return null