PDF parser conforming as much as possible to the PDF specification from Adobe
https://github.com/yob/pdf-reader/