Add field(s) like <metadata|provenance>_<name|schema> to the base Processor class so that every Processor can explicitly identify the specific metadata/provenance record it should peel off the Pipeline for modification (if applicable for that specific Processor).
I prefer using name to schema for identifying specific items -- to me, schema is used to identify / provide hint to the supposed type of the data field, where name is just a simple string identifier not used for other typing / casting hints.
OTOH default values for the fields are obvious if we use schema: "foxden.reader.FoxdenMetadataReader" for metadata, similar for prov.
Add field(s) like
<metadata|provenance>_<name|schema>to the baseProcessorclass so that every Processor can explicitly identify the specific metadata/provenance record it should peel off the Pipeline for modification (if applicable for that specific Processor).I prefer using
nametoschemafor identifying specific items -- to me,schemais used to identify / provide hint to the supposed type of thedatafield, wherenameis just a simple string identifier not used for other typing / casting hints.OTOH default values for the fields are obvious if we use schema: "foxden.reader.FoxdenMetadataReader" for metadata, similar for prov.