Support for PDF Grafting to overlay OCR results

[hackslashX]

Hi,

The OCR feature is awesome since Tesseract results aren't always up to the mark. I was wondering if PDF grafting can be added, i.e. use the bounding boxes from Azure Document or Google Document Engines (LLMs will not work since they don't output it), and overlay it on top of the resulting PDF. This can be challenging since Paperless also performs a lot of PDF processing, so just asking if it's something that is feasible and can be discussed. I'll be also happy to contribute to this feature.

Thanks!

[gardar]

There is a discussion on the topic at #135 and some initial work has been done in #212