Postcode-Search

A Dockerized Flask-restful application to serve postcode GeoJson from a MongoDB backend.

Endpoint:

http://localhost/postcode/eh111at

returns:

{
    "crs": {
        "properties": {
            "name": "urn:ogc:def:crs:EPSG::27700"
        },
        "type": "name"
    },
    "properties": {
        "postcode": "EH111AT"
    },
    "geometry": {
        "coordinates": [
            324079,
            672543
        ],
        "type": "Point"
    },
    "type": "Feature"
}

Requirements

To run the application locally you will need to have installed both Docker and Docker Compose. Compose is a tool for defining and running multi-container Docker applications. This is used to get the Python application, database and HTTP server configured and all talking to each other.

Python scripts have been written in Python 3.5.

Setup

1. Code-Point Open CSV to GeoJson

   To load this data into MongoDB it must first be convert from CSV to GeoJson format.

   1. Download the OS Code-Point Open zipfile and extract the CSV to a sensible location on your machine.
   2. Place the script mongo/csv_to_geojson.pyin the same folder as the extracted CSV files and run python csv_to_gejson.py. This script requires the python-geojson module installed. Once complete there should be a new folder within the CSV directory which contains a GeoJson file for each CSV that exists. e.g. tr.json.

2. Load GeoJson files to MongoDB

   Now we have GeoJson representations for each postcode we are ready to load this data into MongoDB, a NoSQL database. Each file will be loaded as a seperate collection containing a document for each postcode.

   1. Clone this repository on your machine.
   2. In the file docker-compose.yaml, change the path of the mapped volume to a folder on your machine. This allows us to access and run
   3. From a terminal CD into the repo directory and run docker-compose build. This should pull down the various docker images required by the application.
   4. Right now we are only interested in the container named postcodesearch_db. Check this exists with docker ps -a. This command lists all docker containers.
   5. Enter the container with the command docker exec -i -t <container name> /bin/bash and use mongoshell to create a postcode database.
   6. Run the script mongo/file-loop.sh in the same folder as the GeoJson files. The script iterates over all files with a *.json extension and loads them into a collection of the same filename.