looking at the code it seems .osm and .pbf files are extracted using different code. this makes me think we need to go back to using .pbf files in the cucumber tests.