[cafard]

I have once or twice had luck with the Python Camelot package, which you read about at https://camelot-py.readthedocs.io/en/master/ .

But one can burn quantities of time trying to extract useful information from PDFs, with small results. I wish you luck.