Skip to content

Wrap a REST API endpoint with GraphQL in Python

Follow this tutorial to wrap a REST API endpoint with a GraphQL wrapper to make it accessible via a dedicated GraphQL API.

In this tutorial we will use a standalone Ariadne server, which is a Python library for implementing GraphQL servers. It aims to make it easy and enjoyable for developers to create GraphQL APIs by using a schema-first approach, where you define your GraphQL schema using the Schema Definition Language (SDL) and then map your resolvers to the schema.

For the REST API endpoint we will use the City Search API.

The goal of this tutorial is to create a GraphQL API, which will only use the keyword parameter for the query and return only the name parameter in the response.

Pre-requisites

Before you begin, you need to:

  • Register your application with Amadeus for Developers as described in Making your first API call.
  • Have Python installed on your machine.

Create a new Python project

  1. Open your terminal and create a new directory for this project:
    mkdir graphql-wrapper
    
  2. Navigate to the directory:
    cd graphql-wrapper
    

Install required dependencies

Install uvicorn and requests packages by running:

pip install uvicorn
pip install requests

Define GraphQL schema

Create a schema.graphql file with the necessary types and queries. In this tutorial, we are only using the keyword parameter to query the City Search API and we are only interested in the name parameters that this query returns in the response data. For this reason, our schema.graphql will look as follows:

1
2
3
4
5
6
7
type Query {
  getCities(keyword: String!): [City]
}

type City {
  name: String
}

Create a data fetching function

Create a fetch_data.py file and define a function that fetches data from the REST endpoint:

import requests

def fetch_cities(keyword, token):
    endpoint = f"https://test.api.amadeus.com/v1/reference-data/locations/cities?keyword={keyword}"

    headers = {
        "Authorization": f"Bearer {token}"
    }

    response = requests.get(endpoint, headers=headers)
    data = response.json()

    print("API Response:", data)

    return data["data"]

In the above example we are outputting logs to the console for easier troubleshooting.

Implement GraphQL resolvers

Create a resolvers.py file and implement the resolver functions for the queries:

1
2
3
4
5
6
from fetch_data import fetch_cities

def resolve_get_cities(_, info, keyword):
    token = info.context["request"].headers["authorization"]
    cities = fetch_cities(keyword, token)
    return [{"name": city["name"]} for city in cities]

Set up the Ariadne server

Create the main file main.py that sets up the Ariadne server with the schema and resolvers:

from ariadne import QueryType, make_executable_schema, load_schema_from_path
from ariadne.asgi import GraphQL
from resolvers import resolve_get_cities

type_defs = load_schema_from_path("schema.graphql")
query = QueryType()
query.set_field("getCities", resolve_get_cities)

schema = make_executable_schema(type_defs, query)
app = GraphQL(schema, debug=True)

Run the server

Open the terminal and run:

uvicorn main:app

Query the GraphQL API

Information

Before running the query, make sure to obtain the token as described in our Authorization guide.

Now that we have the server running, we can send requests to it. The most straightforward method to do this is by using curl. To query this API by the keyword "Paris":

1
2
3
4
5
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: <your_bearer_token>" \
  -d '{ "query": "{ getCities(keyword: \"PARIS\") { name } }" }' \
  http://localhost:8000/

If your token is valid, the above command will return a list of city names that contain the word Paris.


Last update: April 15, 2024