Whereas the XmlUpdateRequestHandler handles its whole request within a single class, the CSVRequestHandler delegates almost all of its implementation to a CSVLoader class. Is there a short answer to why there would be a split in one case and not the other? (I guess I'm bothered because I'm having trouble coming up with a nice, concise description for myself of the purpose of CSVLoader, distinct from the handler.)
I ask in the context of thinking about *other* updating handlers, e.g. about whether the rich document handler should continue to have its own "loader" class (RichDocumentLoader), or whether it wouldn't be cleaner to move everything to the main RichDocumentRequestHandler class.
