对 POJO (Java) 的所有带注释的字段应用过滤器

Question

Imagine that there are such kind of POJO classes, that just keep data:

public class Pojo() {

  @AnnotatedProp
  String someField;

  SubPojo someSubPojo;

  String someOtherFieldA;

  String someOtherFieldB;
}

public class SubPojo() {

  @AnnotatedProp
  String someSubField;

  Integer someOtherFieldC;
}

someField of Pojo and someSubField of SubPojo are marked special with the @AnnotatedProp property.

I'd like to modify an object of type Pojo . All String fields with @AnnotatedProp annotation should be modified. A "filter" should modify the values of these fields, eg replace some characters inside.

I tried with FieldUtils / simple reflection, but I ended up in stack overflows (the exception AND this forum).

What would be the best way to filter these fields?

Thanks for help.

Answer 1

I've written a recursive POJO filter/transformer doing just that, because I needed it for a project. Just cleaning up and getting green light for release, but the key points are:

• use FieldUtils.readField(field, node, true) for traversal - for getting all fields annotated with your Annotation, there is also a direct method FieldUtils.getFieldsListWithAnnotation() but I needed a more flexible way to scan all nodes first, so I can drill down on sub objects
• you need to check against your own base package name to identify your custom POJO classes and only traverse those sub objects.
• once you have your annotated field, you can simply use FieldUtils.writeField(field, node, getMyValueForField(field), true);

Caveats:

• to avoid a StackOverflow for back-references (child->ancestor), I keep a HashMap and store the FQCN of the nodes I traverse into
• before rolling my own, I checked Apache commons ObjectGraphIterator , but found it not suitable for my purpose.