Allow us either to skip or to filter out the "uninteresting" document part using the event monitoring and then
create memory-based tree from the "interesting" part of a document only and that part process.
