add ocr module to docker

[svettwer]

Acceptance criteria:

• @sakuli/ocr is installed in the docker container

[svettwer]

To process with tesseract > 4.1.1 an update to Ubuntu 20.04 is required.

Now it would be really cool to have a team with whom I could discuss 😅
There are several possible solutions and I'm really not sure which is the best right now.

1. We use snap, the new package manager for Ubunut 20.04 => this requires a lot more HDD space => bigger image (we don't really want that)
2. We only use snap for packages that otherwise no longer work (e.g. chromium) => the container is still bigger
3. We stay on 18.04 and OCR just sucks because of the old tesseract
4. We are starting with a redesign of the container to make it slim, fast and maintainable. Whereby one cannot really guarantee that it will become "maintainable" on the long run.

The only way in which I can guarantee that we will still get deliverable results this week is 3. However, it is questionable how impressive OCR is, if it just works like shit because of the old tesseract version.

All other approaches require some conception. 1 + 2 rather less, 4 more. I'll be there for about a week and then Cliff will seamlessly take over. A handover / coordination etc. is actually not possible. This harbors certain risks as far as the progress of the selected solution approach is concerned.
But I see 1 + 2 critically. Even without the snaps, the container has grown by 500MB again. So a total of 2.65GB now. We should not continue with a growing image at all.