Does iText create image-based or text-based PDF files?

I currently use iText for PDF generation and I am having difficulty determining if the PDF files generated with iText are text-base开发者_Python百科d or image-based once generated. Is there an easy way to determine that programmatically (or to specify one option or the other at the time of generation)?

iText generates PDF text instructions for text, and PDF image XObjects for images. Some other elements (e.g. borders) are generated as PDF graphics instructions. So I suppose you could say it generates "text-based" files.