This metadata generator uses the IPTK web API to watch for new datasets through the /v3/logs/dataset_changes endpoint. If a previously unseen dataset id is detected, the generator looks for DICOM data within the dataset by searching for .dcm files. If a DICOM file is found, all well-known metadata is extracted with the help of pydicom and saved under metadata specification 32bdac29d951d9def51e3cee10c4f0e582f2a962.
Either use the generator.py script directly, which requires Python 3.6 and the pydicom, redis and requests packages, or use the provided Docker image.
The API endpoint can be specified through the API_ENDPOINT environment variable. It must not contain the /v3/ version specification and may contain a username and password to make authenticated calls. A valid endpoint looks like this: http://user:pass@api.server.com. If you pass a Redis host in the REDIS_HOST environment variable, processed dataset ids will be remembered across restarts.