Add support for PDF files

[tempelmann]

PDFs that contain scanned image with text in them can't be parsed with this tool right now because it assumes the files are images.

I wonder if this can be made to work with PDFs as well. I suspect this would require to open the PDF and then render each page into an NSImage, and then feed each image into the text recognizer, and then merge the page results into one.

[xulihang]

This project is for integrating macOS's OCR ability to other full-fledge OCR tools like ImageTrans, so PDF support will not be considered. You can use ImageTrans for PDF OCR.