Add a default/fallback collection area-id in case not available from input messages.
When running the runner on locally received direct readout (DR) data there may not be any geographic gatherer or area-collection processing before. Often the entire local pass is passed on to the runner. In those cases there may not be any collection_area_id parameter in the messages. In those cases the scene registry in the runner would use None for this parameter. That prevents the possibility to filter for redundant EARS/RARS scenes coming from another stream. Fot the EARS-AVHRR there will likely be gatherers prior to aapp-processing and the input messages will indeed have a collection_area_id set to some area of interest. If that area of interest is over the local station there will be many scenes that are in practice the same (covering the same area) - same start and end times and same satellite. Since None will be different from any such area AAPP will not recognice that it had processed the "same" scene just prior (usually the DR data will be a few minutes earlier than EARS).
Add a default/fallback collection area-id in case not available from input messages.
When running the runner on locally received direct readout (DR) data there may not be any geographic gatherer or area-collection processing before. Often the entire local pass is passed on to the runner. In those cases there may not be any
collection_area_idparameter in the messages. In those cases the scene registry in the runner would useNonefor this parameter. That prevents the possibility to filter for redundant EARS/RARS scenes coming from another stream. Fot the EARS-AVHRR there will likely be gatherers prior to aapp-processing and the input messages will indeed have acollection_area_idset to some area of interest. If that area of interest is over the local station there will be many scenes that are in practice the same (covering the same area) - same start and end times and same satellite. SinceNonewill be different from any such area AAPP will not recognice that it had processed the "same" scene just prior (usually the DR data will be a few minutes earlier than EARS).This PR seeks to solve this.
pytestflake8