KOGITO-5534: Codegen should use (un)marshallers for XmlElement(Ref) type #87
Conversation
| @@ -269,11 +269,6 @@ public String getQualifiedName() { | |||
| return null; | |||
| } | |||
|
|
|||
| @Override | |||
There was a problem hiding this comment.
Deleted, it overrode the super class method but did nothing different.
| @@ -149,11 +149,6 @@ public boolean isAttribute() { | |||
| return property.getAnnotation(XmlAttribute.class) != null; | |||
| } | |||
|
|
|||
| @Override | |||
There was a problem hiding this comment.
Deleted, it overrode the super class method but did nothing different.
| @@ -90,7 +90,10 @@ private void processBean(TypeElement bean) { | |||
|
|
|||
| private void processField(VariableElement field) { | |||
| if (checkField(field)) { | |||
| checkTypeAndAdd(field.asType()); | |||
| // Ensure the serializer/deserializer is generated for the fields concrete type, considering | |||
There was a problem hiding this comment.
Don't assume the field's type is correct.. try to look it up from annotations too.
|
@manstis very cool, thanks a million ! |
|
@treblereel Happy to help... it'd be great if you'd be able to release a newer version (with this fix) at some point. |
|
@manstis i think, i can switch jaxb to jakartaee and release all together, wdyt ? |
|
@treblereel You could... but once the I'm more than happy to move my work to Jakarta however IDK if the DMN or BPMN teams would be affected? If they are, then the move to Jakarta probably needs to be coordinated; or at least communicated. |
|
@manstis ok, let's wain for Roger... |
|
@treblereel I'll send an email tomorrow. |
See https://issues.redhat.com/browse/KOGITO-5534
In essence I now check for the presence of
@XmlElementor@XmlElementRefon fields when generating the serializer\deserializer. I also do a similar check when processing the fields on beans to ensure code is also generated for classes not annotated with@XmlMapper.