pdfminer has changed it's API and broken some links

Nobody has claimed this yet.

The euske / pdfminer repository has changed the location of the PDFDocument class, as noted in the README. This class can be refound easily, but also other things have changed as can be deducted from the following error message. I will not pursue this any further and use pdfminer directly.

```
[..]/pdfsplitter/content_extractor/pdfreader/util/convert.py in ()
2 from pdfminer.pdfparser import PDFParser
3 from pdfminer.pdfdocument import PDFDocument
----> 4 from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter, process_pdf
5 from pdfminer.pdfdevice import PDFDevice, TagExtractor
6 from pdfminer.converter import XMLConverter, HTMLConverter, TextConverter

ImportError: cannot import name process_pdf
```

Start with content_extractor/pdfreader/util/convert.py and compare its pdfminer imports with the API changes noted in the README. Trace the reported process_pdf ImportError and identify the other changed imports; done means the converter module loads without these import failures.

Written by the indexing model from the issue text.