Consider handling earlier versions of PDFs

[mlissner]

See: https://github.com/enferex/pdfresurrect

[mlissner]

PyMuPDF doens't suppose this because MuPDF doesn't:

pymupdf/PyMuPDF#3092

I'm chatting in the MuPDF discord to see if they want to add the functionality (or want me to). I don't know C++ though, so I suspect the best thing to do is what Claude suggested and just parse it manually (the MyPDF folks suggested this too).

[mlissner]

I attempted to add this to MuPDF via a PR: ArtifexSoftware/mupdf#88

I've flagged it for them in their Discus chat. Hopefully it'll get picked up.