Where we define all the services for the front-end using a NodeJS and Express.
On a development environment you can use a .env file to map the environmental variables. For the docker container, update the values in the Dockerfile. See Serving static files to map the files folder.
PORT=8000
IMPORTS=imports # import folder
PIPELINE=/location/of/some/python/script/to/run/pipeline
FILES_LOCATION=/mnt/files
The back-end server serves the PDFs and figures to the front-end using the server:port/files/ route. In the docker container, this is configured with the environmental variable FILES_LOCATION that maps to /mnt/files. Therefore, when starting the docker container, pass the parameter -v /route/to/files/:/mnt/files where /route/to/files.
NodeJS can start the curation pipeline by spawning the Python process. However, it does not track if the process ends successfully as the pipeline may process a large batch of documents. We are working on adding monitoring features for the end user. Also, for convenience, the docker container includes the pipeline as part of the back-end image. To use an independent pipeline container, check its corresponding repository.