doc-lit-wrapped WSDLs with wrapper elems with non-substitution-group refs are
incorrectly treated as non-wrapped
----------------------------------------------------------------------------------------------------------------
Key: TUSCANY-1938
URL: https://issues.apache.org/jira/browse/TUSCANY-1938
Project: Tuscany
Issue Type: Bug
Components: Java SCA Core Runtime
Affects Versions: Java-SCA-1.0
Reporter: Scott Kurz
The JAX-WS Spec, Sec. 2.3.1.2, makes the following confusing statement in
defining the WSDL style which qualifies for mapping to "wrapped" Java:
...
(v) The wrapper elements only contain child elements, they must not contain
other structures such as
wildcards (element or attribute), xsd:choice, substitution groups (element
references are not permitted) or attributes; furthermore, they must not be
nillable.
I think the Tuscany interpretation of this statement is wrong in treating
certain WSDLs as "non-wrapped" rather than "wrapped". This makes it hard
to, say, consistently
choose to work in the doc-lit-wrapped WSDL style with a pre-existing WSDL file.
I'll attach a WSDL file for reference, but let me show an excerpt.
This is not treated as wrapped by Tuscany:
<element name="person" type="tns:Person"/>
<complexType name="Person">
<sequence>
<element name="firstName" type="xsd:string"/>
<element name="lastName" type="xsd:string"/>
</sequence>
</complexType>
<element name="getGreetings">
<complexType>
<sequence>
<element name="greeting" type="xsd:string"/>
<element ref="tns:person"/>
</sequence>
</complexType>
</element>
IMO, the JAX-WS RI toolset, (wsimport) has the best interpretation of the
JAX-WS spec. The 'wsimport' tool will treat this as wrapped, (i.e. gen the
Java with @RequestWrapper), which
I think is correct. (BTW, a service impl that I wrote from this WSDL worked
fine on an IBM WAS 6.1 runtime with JAX-WS support.)
In the sample WSDL attached, I show what I think the JAX-WS spec is trying to
disallow. The key point I think is that they're trying to disallow use of
substitution groups. So, if 'person' above
had been an abstract element with an associated substitution group, then JAX-WS
RI would map this in a non--wrapped style.
This also fits with the fact that the statement "(element references are not
permitted)" is not made in a separate sub-bullet, but only in parentheses
directly next to "substitution groups".
Though I still can't explain exactly why the authors phrased it the way they
did !
As further proof this is confusing, here's another JIRA on the exact same
subject on the CXF project (which apparently hasn't been resolved).
https://issues.apache.org/jira/browse/CXF-1079?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12531198
Within the Tuscany runtime, I believe this function is implemented in the
WSDLOperationIntrospectorImpl$Wrapper.getChildElements() method.
There may or may not be a Tuscany tools hit here as well. I haven't checked.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
