You are viewing a plain text version of this content. The canonical link for it is here.
Posted to axis-tck@ws.apache.org by Sam Ruby <ru...@apache.org> on 2002/06/28 21:49:45 UTC
[Fwd: TCK Issue Comments + Some More Issues]
Lance,
I know that you are now subscribed to the mailing lists, and not knowing
if you would prefer getting these directly from me, I am redundantly
forwarding on this e-mail too.
Once Apache gets access to the latest TCK, I will see to it that these
tests are rerun with the latest and will comply with the templates provided.
- Sam Ruby
-------- Original Message --------
Subject: TCK Issue Comments + Some More Issues
Date: Fri, 28 Jun 2002 14:28:43 -0500
From: scheu@us.ibm.com
Reply-To: axis-tck@xml.apache.org
To: axis-tck@xml.apache.org
My comments in <rjs> ...</rjs>
1. Problem: The echoMix method used by the parameterModeTest has the wrong
parameter order! This is causing the TCK to fail for Axis, because Axis
is using the correct parameter order.
This issue was addressed in the final version of the jaxrpc tck
2. Boolean Value Issue
The TCK expects a bean that has a boolean property foo to have a getter
named "getFoo()". This conflicts with the Java Bean Specification. Axis
generates the appropriate "isFoo()" getter.
We have logged bug 4708351. If you could please forward which tests
are failing, that would help us make sure we have addressed all of the
issues.
3. XML to Java mapping problems
One example is the get<wsdl-port-name> method generated in the Service
interface. If the wsdl port name is "foo", the TCK is expecting "getfoo".
Not only does this violate JSR 101 chapter 20, but it is also not
conventional java. Axis generates "getFoo".
We do not have enough information to work through this particular issue.
If you could please provide us with the tests in question and walk us
through the problem you encountered with the test that would be most
helpful.
<rjs>
The failing test is in ..w2j.rpc.encoded.xmlnamemapingtest.Client. The
name of the wsdl port is xMLNameMappingTestService. According to Chap
20, this becomes XMLNameMappingTestService when converted to a java
name. The name of the generated Axis Service operation is
getXMLNameMappingTestService(). The TCK expects
getxMLNameMappingTestService, which results in a test failure. So this
problem is another aspect of the xml<->java name mapping.
</rjs>
4. Service.getPorts issue
com.sun.ts.tests.jaxrpc.api.javax_xml_rpc.Service.Client.GetPortsPosTest2
expects an empty Iterator from service.getPorts(). We throw a
ServiceException because, according to the JAX-RPC javadoc, "Throws:
ServiceException - If this Service class does not have access to the
required WSDL metadata". We believe the TCK should accept a
ServiceException.
We concur that this is an issue and it is being addressed via bug 4709354.
5. Roundtrip issues
There are at least 2 roundtrip issues with
com.sun.ts.tests.jaxrpc.ee.j2w.marshalltest which is a Java-to-WSDL test.
We have logged bug 4708425.
6. wrapped doc/lit
We believe the tests are valid. Please see the example in section 6.4.3
of the jaxrpc specification.
<rjs>
You are looking at the first mapping in section 6.4.3. Read futher in
the same section. "Another possible mapping of the doExample method is
as follows. In this example, the part named body in the DoExample
message is mapped as xsd:complexType:
DoExampleResponse doExample(DoExample body) throws
java.rmi.RemoteException;"
This is the kind of mapping that Axis uses. Also note that the last
paragraph of 6.4.1 indicates that the mapping is specific to a JAX-RPC
implementation and is not portable. So this is not a valid test.
</rjs>
7. Full bean constructor issue
The TCK, in various places, assumes the generated beans have full
constructors. While full constructors on beans are certainly reasonable,
they are not mandated by any spec, so we believe the TCK should not
We agree with your assessment and have logged bug 4708365
8. Stub/Call get/setProperty issue
This has been logged as bug 4708407
<rjs>
More Issues:
9. The jaxrpc.ee.sec tests has direct references to the service location
class (BasicAuthServiceTestService1_Impl) and the URIs. These need to be
changed to use the TSJAXRPCInterface.
10. jaxrpc.ee.j2w.marshalltest does not have MySericeException.class in
the J2WMarshallTest.jar. This class is necessary to generate an
appropriate wsdl file.
</rjs>
Rich 'Shirley' Scheuerle
IBM WebSphere & Axis Web Services Development
512-838-5115 (IBM TL 678-5115)
Re: [Fwd: TCK Issue Comments + Some More Issues]
Posted by Lance Andersen <La...@sun.com>.
Hi Sam,
I have not received any email from the list so I am not sure if it has
been pushed out yet.
Regards,
Lance
Sam Ruby wrote:
> Lance,
>
> I know that you are now subscribed to the mailing lists, and not
> knowing if you would prefer getting these directly from me, I am
> redundantly forwarding on this e-mail too.
>
> Once Apache gets access to the latest TCK, I will see to it that these
> tests are rerun with the latest and will comply with the templates
> provided.
>
> - Sam Ruby
>
> -------- Original Message --------
> Subject: TCK Issue Comments + Some More Issues
> Date: Fri, 28 Jun 2002 14:28:43 -0500
> From: scheu@us.ibm.com
> Reply-To: axis-tck@xml.apache.org
> To: axis-tck@xml.apache.org
>
>
>
> My comments in <rjs> ...</rjs>
>
>
>
> 1. Problem: The echoMix method used by the parameterModeTest has the
> wrong
> parameter order! This is causing the TCK to fail for Axis, because Axis
> is using the correct parameter order.
>
>
> This issue was addressed in the final version of the jaxrpc tck
>
> 2. Boolean Value Issue
>
> The TCK expects a bean that has a boolean property foo to have a getter
> named "getFoo()". This conflicts with the Java Bean Specification. Axis
> generates the appropriate "isFoo()" getter.
>
>
> We have logged bug 4708351. If you could please forward which tests
> are failing, that would help us make sure we have addressed all of the
> issues.
>
>
> 3. XML to Java mapping problems
>
> One example is the get<wsdl-port-name> method generated in the Service
> interface. If the wsdl port name is "foo", the TCK is expecting "getfoo".
> Not only does this violate JSR 101 chapter 20, but it is also not
> conventional java. Axis generates "getFoo".
>
>
> We do not have enough information to work through this particular issue.
> If you could please provide us with the tests in question and walk us
> through the problem you encountered with the test that would be most
> helpful.
>
> <rjs>
>
> The failing test is in ..w2j.rpc.encoded.xmlnamemapingtest.Client. The
> name of the wsdl port is xMLNameMappingTestService. According to Chap
> 20, this becomes XMLNameMappingTestService when converted to a java
> name. The name of the generated Axis Service operation is
> getXMLNameMappingTestService(). The TCK expects
> getxMLNameMappingTestService, which results in a test failure. So this
> problem is another aspect of the xml<->java name mapping.
>
> </rjs>
>
>
>
> 4. Service.getPorts issue
>
> com.sun.ts.tests.jaxrpc.api.javax_xml_rpc.Service.Client.GetPortsPosTest2
> expects an empty Iterator from service.getPorts(). We throw a
> ServiceException because, according to the JAX-RPC javadoc, "Throws:
> ServiceException - If this Service class does not have access to the
> required WSDL metadata". We believe the TCK should accept a
> ServiceException.
>
> We concur that this is an issue and it is being addressed via bug
> 4709354.
>
>
>
> 5. Roundtrip issues
> There are at least 2 roundtrip issues with
> com.sun.ts.tests.jaxrpc.ee.j2w.marshalltest which is a Java-to-WSDL test.
>
> We have logged bug 4708425.
>
> 6. wrapped doc/lit
>
> We believe the tests are valid. Please see the example in section 6.4.3
> of the jaxrpc specification.
>
> <rjs>
>
> You are looking at the first mapping in section 6.4.3. Read futher in
> the same section. "Another possible mapping of the doExample method is
> as follows. In this example, the part named body in the DoExample
> message is mapped as xsd:complexType:
>
> DoExampleResponse doExample(DoExample body) throws
> java.rmi.RemoteException;"
>
> This is the kind of mapping that Axis uses. Also note that the last
> paragraph of 6.4.1 indicates that the mapping is specific to a JAX-RPC
> implementation and is not portable. So this is not a valid test.
>
> </rjs>
I will need to talk further with Rahul and the test leads after our break .
>
>
>
> 7. Full bean constructor issue
> The TCK, in various places, assumes the generated beans have full
> constructors. While full constructors on beans are certainly reasonable,
> they are not mandated by any spec, so we believe the TCK should not
>
>
> We agree with your assessment and have logged bug 4708365
>
>
> 8. Stub/Call get/setProperty issue
>
> This has been logged as bug 4708407
>
> <rjs>
>
> More Issues:
>
> 9. The jaxrpc.ee.sec tests has direct references to the service location
> class (BasicAuthServiceTestService1_Impl) and the URIs. These need to be
> changed to use the TSJAXRPCInterface.
Can you please provide me an example location of this as I do not see
this in the FCS code but I could have missed it.
>
>
> 10. jaxrpc.ee.j2w.marshalltest does not have MySericeException.class in
> the J2WMarshallTest.jar. This class is necessary to generate an
> appropriate wsdl file.
>
> </rjs>
>
This appears to be in the FCS version:
jar tvf J2WMarshallTest.jar
0 Tue May 28 14:35:10 EDT 2002 META-INF/
71 Tue May 28 14:35:10 EDT 2002 META-INF/MANIFEST.MF
3909 Tue May 28 14:32:40 EDT 2002
com/sun/ts/tests/jaxrpc/ee/j2w/marshalltest/MarshallTest.class
11263 Tue May 28 14:32:42 EDT 2002
com/sun/ts/tests/jaxrpc/ee/j2w/marshalltest/MarshallTestImpl.class
569 Tue May 28 14:32:36 EDT 2002
com/sun/ts/tests/jaxrpc/ee/j2w/marshalltest/MyServiceException.class
>
>
> Rich 'Shirley' Scheuerle
> IBM WebSphere & Axis Web Services Development
> 512-838-5115 (IBM TL 678-5115)
>
>
--
Lance Andersen email: Lance.Andersen@sun.com
Sun Microsystems Inc. phone: (781) 442-2037
1 Network Drive, UBUR02-301 fax : (781) 442-1610
Burlington, MA 01803