Currently the tika proposal is about extracting meta information from documents. Now, obviously when dealing with documents this is half the work. Updating/changing meta data of documents is the other half.
When it comes to an implementation, you might already have more than half of the code when you're able to extract meta data. I think it makes sense to extend the scope of Tika to include writing/changing documents as well. We could make this feature optional though. WDYT? Carsten
