This change speeds up OCR extraction by allowing bypassing OCR for texts that are irrelevant (not in table). --------- Co-authored-by: Nguyen Trung Duc (john) <trungduc1992@gmail.com>