A customer has a use case where they need to search for documents and OCR'd text, and be able to export the filenames along with their respective context hits. Simply put, is this even possible?