You can replicate the two dimensional neighborhood by mapping the
resursive partitioning of the lat-lon range into key components.
Eg: lets say you start with north south axis first, if your location is
north of equator, add the first caharacter n to the key, else add s.
Next check if your location is east or west of equator. If east, then
emit e, else emit west.
At this first stage of recursion, you have narrowed down from entire
globe to one quadrant. Now consider the center of this quadrant and
emit a n or s depending on whether your location is north or south or
it, and then e or w depending on whether it is east or west. Now you
have location within 1/16 of the globe.
Keep repeating to get accuracy as desired.
For rxample :
neswnesenwsesesenwswswnwsenwswnwsenwnwnesenwnese would be the code for
(28.5355161 N , 77.39102649999995 E )
Because nearby locations would share the most significant places in the
key, they can be traversed and located using some location prefix.
On 12/8/13, 10:55 AM, Robert Newson wrote:
CouchDB views are one dimensional so you will not succeed with a two
dimensional geo query. You could try couchdb-lucene which can.
On 8 Dec 2013 15:51, "Qaqabincs" <[email protected]> wrote:
I use a view to query an area, and emit [lng, lat] as key, so I use
"...?startkey=[min_lng, min_lat]&endkey=[max_lng, max_lat]" to help find
out all place within the quadranglar area, but the results are exactly
between x?[min_lng, max_lng]&y?[-?, +? ] and x?[-?, +?]&y?[min_lat,
max_lat] (like a cross-shaped stripe area).
what is the correct usage of startkey/endkey to express a quadranglar area?
-----------------------------
qaqabincs
