Introduce DocumentState for FIPSCertificate class

[adamjanovsky]

The DocumentState class is stored for various documents associated with CCCertificate class. The same class should be used to hold data in FIPSCertificate instances

[J08nY]

Connected to this, there is quite a bit of deduplication possible. See what I have done in 5e983ef. Essentially, all of the boilerplate functions on the certificate objects that get parallelized: download_pdf, convert_pdf, extract_pdf_meta, extract_keywords are essentially the same for CC and PP and the only thing that changes is the: link, DocumentState object they store the results to and PdfData object they store the filename to. With FIPS unifying to use DocumentState we could simplify this greatly.

[adamjanovsky]

Yes, I've seen that while copying the new functionality for PPs straight from the CC implementation. I didn't want to handle the abstraction at that time. Essentially, you just want to specify the action (download, convert, extract), the source attribute and the target attribute. Then you just execute.

[J08nY]

The thing is that apart from DocumentState the methods set the "*_filename" attribute of PdfData. So basically there needs to be consistency between those two classes, and there is nothing enforcing it (be it typing or something at runtime) if we move to the getattr/setattr implementation.