Simple GraphQL Profile Server

This example is a simple Flask Python implementation of a server that updates from the Murmurations directory and serves up the results using a GraphQL API.

What it does

The server consists of a simple api and cache of profile data (murmurations). The cache is loaded by making a request to murmurations index server, taking the profiles that are available and caching them in a local database. The cached profiles are then exposed through a GraphQl query API.

The GraphQL API gives developers a lot of flexibility over what data they want, how to shape the data when it is returned and many other cool things.

More information about the query and mutation mechanisms take a look at https://graphql.org/learn/

The data is cached in a SqlLite database for simplicity, however, this can be swapped out for another RDBMS in the configuration. See Flask documentation for details.

Note this implementation is intended as an example. To run it in a production configuration it will be necessary to change the configurations to turn off debugging etc.

Project Setup

To run the server locally the easiest thing to do is to run it in Docker.

Ensure that docker v2.x and above is installed

Then run the following commands to build the image and run the container:

$ cd /examples/python
$ docker build -t flask-murmuration .
$ docker run -d  -p 5000:5000 flask-murmuration

When the container is running you can navigate to http://localhost:5000/graphql and interact with the query browser. To load the data from the murmurations idex server you will need to run the update cache mutation. The query looks like this:

mutation{
  updateCache {
    ok
    message
  }
}

When that query has successfully executed the local cache will be populated. You can start to build queries using the query objects. See GraphQL documentation for details.

A basic query looks like this (use the control + spacebar to get additional field suggestions):

query{
  murmuration {
    id
    name
    url
    tagline
    updated
    mission
    location
  }
}

Todo

There are a lot of ways this could be enhanced...

• Update cache automation based on time config
• Data mapping rules and configuration
• CRUD functions for profiles
• Other good ideas...