The CAP alerts file we're parsing contains alerts for Binnensee and Küste warning cells, and Bright Sky correctly parses these and returns them when calling /alerts without any parameters (as can often be seen in the alerts demo).
However, it is currently not possible to directly query for alerts for these cells directly (i.e. to supply their cell ID as warn_cell_id argument), b/c we only load meta data on the Gemeinde cells.
While meta data for the Binnenseen and Küsten warn cells is easily available from the DWD's GeoServer, we cannot just mix them in with the Gemeinde cell data, as the Binnensee cells have overlapping polygons with the Gemeinde cells and would screw with selecting cells by lat/lon. I'm currently leaning towards storing the Binnenseen/Küsten cell meta information in an extra dict that serves as fallback if a given warn_cell_id can not be found in our regular meta dict. That would allow retrieving warnings for Binnenseen/Küsten by cell ID but not by lat/lon.
The CAP alerts file we're parsing contains alerts for Binnensee and Küste warning cells, and Bright Sky correctly parses these and returns them when calling
/alertswithout any parameters (as can often be seen in the alerts demo).However, it is currently not possible to directly query for alerts for these cells directly (i.e. to supply their cell ID as
warn_cell_idargument), b/c we only load meta data on the Gemeinde cells.While meta data for the Binnenseen and Küsten warn cells is easily available from the DWD's GeoServer, we cannot just mix them in with the Gemeinde cell data, as the Binnensee cells have overlapping polygons with the Gemeinde cells and would screw with selecting cells by lat/lon. I'm currently leaning towards storing the Binnenseen/Küsten cell meta information in an extra dict that serves as fallback if a given
warn_cell_idcan not be found in our regular meta dict. That would allow retrieving warnings for Binnenseen/Küsten by cell ID but not by lat/lon.