Ah, I see. Thanks for sharing. I am OCR'ing public domain .pdfs, so perhaps I haven't run into that issue. The irony couldn't be stronger though.