If you mean that the missing schemaLocation should be flagged: yes, create a
(minor priority) issue.
-W
On Sun, Mar 29, 2009 at 7:11 PM, Benson Margulies <bimargulies_at_gmail.com>wrote:
> Wolfgang,
>
> I think I agree. Someone on the CXF project thought it would be clever to
> interpret this as 'any schema in the WSDL file of the moment', on the theory
> that the XPath could include the schema element itself. This strikes me as
> 'do what I mean.'
>
> Does this net out to a small JAXB defect involving reporting an error?
>
> -benson
>
>
> On Sun, Mar 29, 2009 at 1:00 PM, Wolfgang Laun <wolfgang.laun_at_gmail.com>wrote:
>
>> I would agree with you if you'd say that an XPATH reference should cause
>> an error if
>> it isn't in the scope of some <xjb:binding schemaLocation="xxx"> element.
>> Customizations
>> don't make sense if directed to binding file elements.
>>
>> But searching through any number of schemas for a "lucky hit" doesn't
>> appear
>> quite attractive, either. What defines the order? What if it matches in
>> A.xsd while
>> you think it should match in B.xsd?
>>
>> -W
>>
>>
>> On Fri, Mar 27, 2009 at 4:28 AM, Benson Margulies <bimargulies_at_gmail.com>wrote:
>>
>>> Consider a binding file with no schemaLocation attribute.
>>>
>>> com.sun.tools.xjc.reader.internalizer.Internalizer.buildTargetNodeMap(Element,
>>> Node, Target, Map<Element, Node>, SCDBasedBindingSet)
>>>
>>> will apply the xpath to the binding file itself (the 'inherited target').
>>> Should it not apply it to each of the schemas known to the compilation,
>>> instead?
>>>
>>>
>>>
>>
>