Getting the markup inside an entity

Hi all.

I am writing some xml based import/export features for GNU
Ferret[1]. The data i want to import/export is somewhat like:

<PROJECT>

  <DATAMODEL NAME="..." ...>
    <OBJECT_COLLECTION>

        <OBJECT OID="..." ...>

          <DATA>
            <!-- Object specific data -->
          </DATA>

        </OBJECT>
  
        ...

    </OBJECT_COLLECTION>
  </DATAMODEL>
  
  <DOMAIN_TREE>

    <!-- Domain tree data -->

  </DOMAIN_TREE>

  ...
</PROJECT>

Each object in a datamodel has a type. Each object type store a
different data structure. GNU Ferret is written in a way in that the
developer can "register" new object types. 

I want to use different expat parsers to manage the <DATA> entities
that contain the data of an object. In the same way i want to use a
different parser to manage the domain trees (the user may want to
import a domain tree stored in a separated file, beginning with the
<DOMAIN_TREE> entity).

Well, as i understand it, what i need is a way to get the complete
markup inside an entity. For example, when the datamodel parser finds
a <DATA> entity, it would get the markup inside that entity, call the
object type specific parser to process it, and ignore the contents of
<DATA> continuing the parsing.

Is that feasible? If not, there is another way to achieve the
modularity i want? (i was thinking in complex temporal handler and
user data reassignments).

Thanks in advance!!


[1] - http://www.gnu.org/software/ferret