See https://github.com/westnordost/countryboundaries
It's per node. Size and supported admin levels depend on the data you stick in there. (Also, the faster the access should be, the bigger the file.) The default data is a geoJson file from the JOSM project which has all the countries, and states only for the United States, Canada, Australia, China and India and is about 1MB. > I imagine the same calculation could be used for other default > calculations, like the default language format, Here is a list of official languages by country iso code: https://github.com/westnordost/StreetComplete/blob/master/res/country_metadata/officialLanguages.yml Cheers Tobias On 20/09/2018 00:07, Joseph Eisenberg wrote: > I’m encouraged by this: > > > >”3. removes superfluous country ISO-code. Country boundaries > information > is already available in the OSM > database, data users can just use a > (reverse) geocoder to find where a street is. > (I wrote Java library recently that does this offline in ~0.1ms)” > > > Does this offline reverse geo coding calculation require downloading the > global admin_level=2 boundary data? Is it a reasonable amount of data? > > Does it take 0.1ms per node, or can several highway segments be checked > in the same calculation? > > I imagine the same calculation could be used for other default > calculations, like the default language format, > > Or it could be used to display certain features differently in different > countries: e.g. don’t use a cross symbol for hospitals in Saudi Arabia. > > For the USA, how quickly can you reverse geocode admin_levels 2, 4, 6 > and 8 for a road? (National, State, County and Municipal) > > > > > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging > _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
