Validator failing when processContents is skip, but unresolvable xsi:type is present

[GoogleCodeExporter]

If a wildcard's processContents is set to skip, but a child has an xsi:type 
attribute that's unresolveable, we're raising an exception for the unresolved 
xsi:type.  We should not.

Original issue reported on code.google.com by joe.bays...@gmail.com on 8 Apr 2014 at 7:09

[GoogleCodeExporter]

r503.  Ignore unresolved xsi:types if processContents is skip or lax.

Original comment by joe.bays...@gmail.com on 8 Apr 2014 at 7:12

• Changed state: Fixed

[GoogleCodeExporter]

Re-opening this ticket because we're still failing on the following use case:  
element w/unresolveable xsi:type attribute matches a wildcard (in its parent 
content model) whose processContents is set to skip or lax.  In that case, 
we're still incorrectly reporting an error. 

Original comment by joe.bays...@gmail.com on 23 Sep 2014 at 6:21

• Changed state: Started

[GoogleCodeExporter]

r517.  AttributeManager.initialize(...) now returns a QName for unresolved 
xsi:types instead of a boolean.  That QName is non-null only if an xsi:type is 
not resolveable.  ValidationKernel.startElement now reports an error if 
AttributeManager.initialize returns a QName (see above) and the current 
processContents is strict.

Original comment by joe.bays...@gmail.com on 23 Sep 2014 at 7:55

• Changed state: Fixed