This extracts tables from PDFs. It supports cells spanning multiple rows or columns - View it on GitHub