Fig. 2

Implementation of an Archivist metadata processing pipeline with two example parsers. From the user perspective, the Archivist class is provided with all inputs associated with the pipeline (bottom yellow and grey boxes). Internally, the Explorer class extracts the individual files to process from a collection of raw metadata files, and dispatches them to corresponding Parser classes, here the Configuration parser and the Time parser classes. Then, each Parser class employs a user-defined function to extract specific information from its respective files. After this, the Formatter class collects the parsing results. If a data template is provided, the composite result can also be restructured. The final processed metadata is output in a format of choice by the Exporter class. A Jupyter tutorial explaining how to implement the metadata processing pipeline shown here is available at https://github.com/INM-6/metadata-archivist/blob/main/examples/schema_example4/schema_tutorial.ipynb (last access: 26 January 2025).