This Docker image is based on the official Nextcloud image but additionally installs OCRmyPDF to allow the usage of the OCR workflow in Docker based installations. There are no further changes to the image.
All the languages result in a rather large image. Maybe at some point, I'll split it up into different images to keep them small.
There is a daily GitHub action which pulls the latest tag, creates the corresponding new images, and pushes them to Docker Hub. The images start with version 24.0.3 and include the -apache and the -fpm image. If a new version is missing, it should be available tomorrow ;)
For a quick test, just run docker run -p 8080:80 --rm mjdazz/nextcloud-ocrmypdf:latest, access the instance via your browser under http://localhost:8080 and install the OCR workflow. Everything should work as expected
If you're running docker compose, remember to replace the image of the container running the cron task!
If you're running Nextcloud on K8s with the official Helm Chart, you only need to replace the image of the app, not of the cron job, as it uses WebCron:
image:
repository: mjdazz/nextcloud-ocrmypdf
flavor: apache
# If unset, uses flavor + .Chart.AppVersion to create tag
tag: 24.0.3
pullPolicy: IfNotPresent