Hi Nadir, Thanks for the fix you made to the test case. About the deserialization problem:
I investigated about the problem and found that the deserializer cannot correctly handle the wrapper element of the array in RPC style SOAP messages. I guess that the problem is occurred when the deserializer is trying to deserilize the following message. <ns1:echoStringArray xmlns:ns1="http://soapinterop.org/";> <inputStringArray xmlns:enc="http://www.w3.org/2001/06/soap-encoding"; enc:arrayType="xsd:string[2]"> <item>0th element of string array</item> <item>1th element of string array</item> </inputStringArray> </ns1:echoStringArray> To get the array size from the above message, int SoapDeSerializer::getArraySize method is called. I guess that this is the place where it all messes up. It correctly returns the array size. But after that, it fails to get the array content in item tags. Hope this information will be helpful to you. Regards, Dushshantha -----Original Message----- From: Nadir Amra [mailto:[EMAIL PROTECTED] Sent: Saturday, December 24, 2005 4:24 AM To: Apache AXIS C Developers List Subject: new memory model problems with testcases Hi Dushshantha, Thanks for fixing. In InteropTestRound1Client.cpp, I corrected the SOAPStruct and added further checks to check for returned array sizes before assuming success. For example: xsd__string_Array arrstr; xsd__string* sToSend = new xsd__string[ARRAYSIZE]; sprintf(buffer1, "%dth element of string array", 0); sprintf(buffer2, "%dst element of string array", 1); sToSend[0]= buffer1; sToSend[1]= buffer2; arrstr.set(sToSend, ARRAYSIZE); ws.setTransportProperty("SOAPAction" , "InteropBase#echoStringArray"); printf("invoking echoStringArray...\n"); int outputSize = 0; xsd__string_Array* outPutStrArray = ws.echoStringArray(&arrstr); if (outPutStrArray != NULL && outPutStrArray->get(outputSize) != NULL && outputSize == ARRAYSIZE) printf("successful\n"); else printf("failed: outputsize=%d\n", outputSize); Assuming my if-check above is correct, the array size checks fail (always zero). I am assuming that is the problem you were mentioning with deserialization? Andrew will have to look at this further although I will continue to see if I can resolve. Nadir K. Amra "Dushshantha Chandradasa" <[EMAIL PROTECTED]> wrote on 12/23/2005 01:11:23 AM: > Hi Nadir, > The problem is with the Client implementation. It was not suit to the > new memory model. I fixed some of these problems and committed the > changes. I commented out the SOAPStruct array part in the file. Have to > correct that as well. > > Regards, > Dushshantha >
