OCR on demand should offer more than a single OCR pipeline.
Users can select the desired pipeline by its name. That name selects a script / program which is called with the required parameters (METS URL, page number).
Maybe: Use JSON file for mapping between script name and offered pipeline name.
OCR on demand should offer more than a single OCR pipeline.
Users can select the desired pipeline by its name. That name selects a script / program which is called with the required parameters (METS URL, page number).
Maybe: Use JSON file for mapping between script name and offered pipeline name.