This project is a homework assignment for the Geodesy and Geominformation Master program at the TU Berlin. The idea is to allow a user to find Apartments given some additional information on a map. This is a rough prototype and only demo features are enabled. It is meant to work for buildings in Berlin. Since this is just a fast prototype, its adapted to work in Firefox and you need to serve it through a web server, such as an Apache.
The following map features are implemented
- Two base maps are implemented. For this purpose only the OSM base map gives a reasonable resolution.
- One thematic WMS layer representing population density.
- Two WFS feature layers:
- One point layer displaying public schools.
- One polygon layer of emissions.
- Address search using the Photon API.
Houses are found by applying filters, since the data of pulling all addresses in Berlin would overload the system and not be informative. Filters can be applied using the following tools:
- Creating a specific filter where buildings:
- Have a distance to a certain point feature with a certain property.
- Have a certain type of value (not implemented).
- Are in a certain district or other polygon with a certain type (not implemented).
- Clicking a feature and selecting to search around or in it.
- Double clicking anywhere on the map.
The user can decide to store data in the form of cookies. Additionally, the locations stored can be exported as a GeoJSON.
This is not an optimal solution and a NoSQL implementation of sorts would be a more elegant approach. But the parameters of the task state minimal server configuration and a full CouchDB implementation (or the like) would exceed the scope.
To run and install the package you will need NodeJS installed on your computer. Once you have cloned the repo, run:
npm install
This will install all the dependencies. To build the uglified bundle run:
npm run release
I had some issues with npm and browserify/uglify. If the command does not run, install those two globally
npm install -g browserify
If you care to do some development, you can use the grunt file provided:
grunt dev
grunt watch
To serve everything:
node server.js
The application can now be called from your localhost on localhost:3000.