Is stanley-ai telling me the truth, that documentCapture module's documentToText() function can't do actual OCR, and can only extract text that's already embedded in the PDFs? Well it seems to be so. However, I was able to extract text from such documents using documentToStructure() instead.