|
Hi - I am trying to find the best way to search for "All butchers in England" or "All butchers in Germany". I notice that when I do a search for a butcher via Nominatim in Geopy where I already know the coordinates, Nominatim indeed seems to know it is a butcher: location = geolocator.reverse("50.8907671, 12.4256231") locaton.raw {'place_id': ...other stuff. 'address': {'butcher': 'Fleischerei Hanns', 'house_number': '6', 'road': 'Am Löschkenberg'... If I try an unstructured search I get very few results (should be in the thousands)
And I know you can do structured searches in Nominatim. So how would I structure a search here to return all the addresses that Nominatim thinks is a butcher? PS - I do the same search on Overpass-turbo and get around 5900 entries, but the addresses do not routinely have a CITY and that's what I'm looking for. They only have what people might have entered. I can run the Overpass lat/long through Nominatim reverse search to get the city information, which works but it feels inefficient. If I can answer my first question I'll have a solution. Thanks for any help! |
|
Nominatim can only find POIs (shop, restaurant, hospital) in a short radius. Even then the list can be incomplete. Overpass or filtering a *.osm file with other tools is the better systematic approach for a whole country. Nominatim support batch requests with http://nominatim.org/release-docs/latest/api/Lookup/ The limit is set to 50 on nominatim.osm.org (configurable if you have your own installation). It might seem inefficient but the only option would be direct database access to a Nominatim installation. Super, thanks for the explanation.
(02 Jul '19, 07:45)
matthpau
|
