Issue 1635: Decypting PDF ebboks fixed

[frickeo]

The decryption of pdf ebooks was broken (see #1635).

If fixed this with two modifications

First of all, I got an empty list of objids when parsing a pdf document: In Python 2, the build-in function zip() returns a list and not an iterator. This changed in Python 3. I converted the iterator into a list again and got objids back.

Another correction was needed. to make the predictor code work again.

[abort-retry-fail-ignore]

Works for me. Tried with two problem pdfs, decrypted successfully. Thanks for tracking this down!

[gschenck]

Works. Thank you!

[Leseratte10]

I've found a PDF where this bugfix wasn't enough to make the DeDRM work again.

There's code in the nunpack function that looks like this:

return struct.unpack('>L', b'\x00'+s)[0]

and that gave me some Python error about being unable to combine str and bytes.
Changing that line to the following fixes this issue:

return struct.unpack('>L', bytes([0]) + s)[0]

Maybe that can be added to this PR as well.