Is there a way to get results for an overpass query paginated?

Question

Let's say I want to get restaurants in Berlin and I have this query:

[out:json];
area["boundary"="administrative"]["name"="Berlin"] -> .a;
(
    node(area.a)["amenity"="restaurant"];
); out center;

Let's say this result set is too big to extract in just one request to overpass. I would like to be able to use something like SQL's OFFSET and LIMIT arguments to get the first 100 results (0-99), process them, then get the next 100 (100-199) and so on.

I can't find an option to do that in the API, is it possible at all? If not, how should I query my data to get it divided into smaller sets?

I know I can increase the memory limit or the timeout, but this still leaves me handling one massive request instead on n small ones, which is how I would like to do it.

Answer 1

OFFSET is not supported by Overpass API, but you can limit the number of result this is getting returned by the query via an additional parameter in the out statement. The following example would return only 100 restaurants in Berlin:

[out:json];
area["boundary"="administrative"]["name"="Berlin"] -> .a;
(
    node(area.a)["amenity"="restaurant"];
); out center 100;

One approach to limit the overall data volume could be to count the number of objects in a bounding box, and if that number is too large, split the bounding box in 4 parts. counting is supported via out count;. Once the number of objects is feasible, just use out; to get some results.

node({{bbox}})["amenity"="restaurant"];
out count;