Cannot use on password protected files

[heroic]

Most of the files that we have, have a password. PdfFileReader does support decrypting the file, but without an option to pass the password in read_pdf it cannot be done

[rbares]

There are two issues with this:

• PdfFileReader from PyPDF2 only provides limited support for encrypted PDFs (see PyPDF2 can't decrypt PDF files with Acrobat 6.0 or higher password security compatibility py-pdf/pypdf#378 and Implement Encryption Version 4 py-pdf/pypdf#226)
• PyPDF2 decryption handles some magic values strangely. From my testing unicode passwords in Python 2.7 cause a UnicodeDecodeError and are not currently supported. May be able to resolve this with an upstream PR but the project is largely unmaintained.

I have a potential pull request for camelot to add the mediocre level of support offered by PyPDF2, but am uncertain whether that approach is desirable.

@vinayak-mehta: given the lack of maintenance have you considered moving to an alternative PDF library? PikePDF looks promising but would require backporting for python 2.7

[vinayak-mehta]

@rbares Thanks for the detailed comment! I've faced issues with PyPDF2 decryption in the past (I remember the same problem of it not being able to support some encryption types). I've found qpdf for those cases. I see that PikePDF is based on qpdf, so I could look into it later. But having a password kwarg in read_pdf for PyPDF2's limited encryption support sounds like a good idea, till we move on to something better. You can go ahead and open the PR!

@heroic Maybe you can add a preprocessing step that decrypts all your PDFs at once with qpdf and a wildcard (*.pdf), before extracting tables using Camelot?

[Fabian1337]

any update on this ?
i am not able to move with my system to qpdf so i am looking for some python based solution :/

[vinayak-mehta]

@Fabian1337 Does the the password kwarg introduced in 429640f not work for you? Which system are you on? Is it possible to upload the PDF that you're trying to parse?