Code example for usage of eynollah with docker

[alexander-winkler]

Hello! I might have to run eynollah on MacOs, so I had a look at the docker image. Do you happen to have the commands needed to get the container running and actually use eynollah within the container? Would you mind sharing?

Thanks!

[cneud]

Hi @alexander-winkler and apologies for the late reply! We are currently busy with the preparation of a major new release while also moving offices...but I hope we can get back to you with information about usage of Eynollah in Docker end of next week.

[alexander-winkler]

very kind! no worries, though, it's not terribly urgent.

[bertsky]

Note: you are out of luck with OCR-D Docker images on MacOS. We base ocrd/core (used as basic build stage here) on ubuntu:20.04, which in principle is multi-platform, including arm64/aarch64 (which runs on the Apple M1). But getting OpenCV to build is too much pain for us (I tried in OCR-D/core#1267, but had to give up ultimately). Perhaps in the future, things with scikit-build / cmake-python / opencv-python-headless improve to the point that we can at least try again...

However, getting a compatible version of Tensorflow might be even more of a problem. I already forgot about the details, but TF binaries need AVX instructions, which that processor does not support. And Rosetta (the native / processor-level x86 emulator) does not, either. And IIRC QEMU, Docker's runtime emulator, used to not support it either. (But that seems to have changed recently...)

[bertsky]

But if you feel like trying it out yourself, checkout #151 and run make docker to build ocrd/eynollah yourself, or use docker pull bertsky/eynollah for something equivalent.

The usage is:

# 1. ocrd resource manager
# (just once, to get the models and install them into a named volume for later re-use)
vol_models=ocrd-resources:/usr/local/share/ocrd-resources
docker run --rm -v $vol_models ocrd/eynollah ocrd resmgr download ocrd-eynollah-segment default
# Now, each time you want to use Eynollah,
# pass the same resources volume again.
# Also, bind-mount some data directory, e.g. current working directory $PWD
# (/data is default working directory in the container).
# Either use standalone CLI (2) or OCR-D CLI (3):
# 2. standalone CLI (follow self-help, cf. readme)
docker run --rm -v $vol_models -v $PWD:/data ocrd/eynollah eynollah binarization --help
docker run --rm -v $vol_models -v $PWD:/data ocrd/eynollah eynollah layout --help
docker run --rm -v $vol_models -v $PWD:/data ocrd/eynollah eynollah ocr --help
# 3. OCR-D CLI (follow self-help, cf. readme and https://ocr-d.de/en/spec/cli)
docker run --rm -v $vol_models -v $PWD:/data ocrd/eynollah ocrd-eynollah-segment -h
docker run --rm -v $vol_models -v $PWD:/data ocrd/eynollah ocrd-sbb-binarize -h
# Alternatively, just "log in" to the container once and use the commands there:
docker run --rm -v $vol_models -v $PWD:/data -it ocrd/eynollah bash