Skip to content
Experimental layer management functionality for GeoCam Share
Python JavaScript
Find file
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


1st paragraph describes point 2nd paragraph "here is the exact install instructions"

The point of the geocamLayer project is to provide a server-side application that performs the tasks of managing & clustering points on a map, reducing to a minimum the client-side requirements in the area of processing power. This application provides a json and a kml interface, which both account for clustering automatically. The code itself is designed to be easy to integrate into existing projects, as the view method that generates the clusters is completely autonomous, and includes no database interactions. Documentation of the exact format is included below.

To get the included demo running:

  1. Install django
  2. cd into example
  3. python syncdb
  4. python runserver

The server should now be running on localhost:8000. The web page should be a full-page google map, which should automatically load points and clusters as you move the map around. Clicking on a cluster should bring you down to a zoom level that includes all the points in the cluster.

@section{Technical details}

The "get" function takes the request as a first argument and the list of points as the second. It has two encoding options, geojson and kml.

The geojson is formatted as follows:

  "type": "FeatureCollection",
    {"geometry": {
          (lat-coordinate of feature),
        (lng-coordinate of feature)
    "properties": {
      (for clusters)
      "bbox": [
        (north coordinate of bounding box containing points in cluster),
      (east coordinate),
      (south coordinate),
      (west coordinate)
      (for clusters:)
      "numpoints":(number of points in cluster),
      (for other points:)
      "timestamp":(timestamp of point),
      "timespan":(timespan of point),
      "name":(name of point),
      "description":(description of point),
      (all other properties that are provided by the getProperties function)
    (all the other points on the map)

The KML is formatted as follows:

TODO: this section.

A sample server-side implementation of this could be:

url_patterns += [
    (r'^points/', 'views.get', {'objects':objects})

The values passed to this "get" views are passed in the URL ("GET" parameters) and are as follows (all are optional):

cluster: 1 or 0 for yes or no, respectively bbox: toUrlValue() of a LatLngBounds gmap2 object encoding: geojson or kml start: start index end: end index reverse: 1 or 0 for yes or no, respectively

A sample URL could be:,-142.99263,46.496068,-101.13472

The client side interface for the geojson is designed to be easy to read for gmap clients. The way that it is designed to be implemented is json parsing through an XMLHTTPRequest, which then pushes points or clusters depending on the subtype to the gmap interface. An example implementation is provided in the static folder.

The positions and bounding boxes of the objects sent through the geojson are intended to work well with google maps, and are in the right order to enable direct loading of the points into the gmap. The bounds included in each cluster can be translated into LatLngBounds with no change in order, which can then be passed to the fitBounds method which will correctly zoom down to a point where you can see all the points in the cluster.

Something went wrong with that request. Please try again.