Setup Expo with FaunaDB via Vercel

This post was originally published on my garden.richardhaines.dev

Im currently working on a project which will be a mobile app for parents to swap
unused items. To do this im using Expo with a
FaunaDB database. I had already setup a project with Fauna
and Gatsby so thought the knowledge transfer would be quite simple. And it was,
until it wasn't.

The problem with trying to access the Fauna database from a React Native (RN)
app is that RN doesn't come with or support the Node Standard Library, which
Fauna uses. The solution, or the one I found, is to create a separate API to
handle the Fauna side of things and access that API form the app. Let's go
through a simple setup to see how this is done.

Make sure you have installed the expo-cli and now-cli.

yarn global add expo-cli
yarn global add vercel

Create a new expo project, follow the instructions (you can choose a blank or
tabs template) and open your code editor

expo init myProject
cd myProject
code .

Ok, leave the app for a minute and login to your Fauna account, create a new
database and name it whatever you like. Perhaps MySuperDB? Create a new key with
a server role. Name it FAUNA_SERVER. Make a note of it as it wont be shown
again.

Back to our app we now need to add some packages. Its a long list so just do it
all in one.

yarn add @apollo/client apollo-link-context apollo-link-http graphql graphql-tag isomorphic-fetch

yarn add react-native-dotenv -D

First we want to setup our Apollo Provider which will wrap our root component,
App.js. Create a new file in the constants folder and name it apollo.js

const React = require("react");
import { FAUNA_SERVER } from "react-native-dotenv";
const {
  ApolloProvider,
  ApolloClient,
  InMemoryCache
} = require("@apollo/client");
const { setContext } = require("apollo-link-context");
const { createHttpLink } = require("apollo-link-http");
const fetch = require("isomorphic-fetch");

const httpLink = createHttpLink({
  uri: "https://graphql.fauna.com/graphql",
  fetch
});

const authLink = setContext((_, { headers }) => {
  return {
    headers: {
      ...headers,
      authorization: `Bearer ${FAUNA_SERVER}`
    }
  };
});

const client = new ApolloClient({
  link: authLink.concat(httpLink),
  cache: new InMemoryCache()
});

const Apollo = ({ children }) => (
  <ApolloProvider client={client}>{children}</ApolloProvider>
);

export default Apollo;

The above code can be boiled down to:

• Created a new http link
• Create an auth link which returns the headers to the context so the http link can read them. Here we pass in our Fauna key with server rights.
• Then we create the client to be passed to the provider with the link now set as the auth link.

Create a new .env file at the projects root and add your fauna server key, using
the name we gave it in our fauna dashboard.

FAUNA_SERVER=xxxxxxxxxxxxxxxxxxxxxxxxxxx

Next head over to App.js and import and wrap the App with our new Apollo
component. (I had used the tab template so yours will look a bit different if
you chose the blank templete)

import Apollo from "./constants/apollo";
//.....lots of code

return (
  <Apollo>
    <View style={styles.container}>
      {Platform.OS === "ios" && <StatusBar barStyle="dark-content" />}
      <NavigationContainer linking={LinkingConfiguration}>
        <Stack.Navigator>
          <Stack.Screen name="Root" component={BottomTabNavigator} />
        </Stack.Navigator>
      </NavigationContainer>
    </View>
  </Apollo>
);

Now we can add a schema, which we will import into our Fauna database via their
dashboard. Inside the constants folder create a new file named schema.gql.

type Query {
  allUsers: [User]
}

type User {
  name: String!
  email: String!
}

Back on the Fauna dashboard click the graphql tab on the left and import the
schema. Once done you will be presented with a graphql playground. Add some
users to our database.

mutation {
    createUser(
      data: {
        name: "Richard Haines"
        email: "hello@richardhaines.dev"
      }
    ) {
      _id
      name
      email
    }
  }

Ok cool, now we can setup our serverless function. Create a new project and
initialize it.

mkdir mySuperAPI
cd mySuperAPI
yarn init -y

Vercel expects us to have an api folder so create that. We will also want to
install some packages here.

yarn add @now/node apollo-link-http apollo-server-micro dotenv graphql graphql-tag isomorphic-fetch

Inside the api folder create a new folder that begins with an underscore, this
is so that vercel knows not to build it's content as a functions. Inside that
folder create a new file named config.js.

cd api
mkdir _utils
cd _utils && touch config.js

Create a .env file at the root and add the Fauna server key. Then in the new
config.js file add the following. This is basically a short cut for us having to
type out process.env.MY_KEY. Right now we only have one key but we might want to
add more later. I find this a handy little file to use.

const dotenv = require("dotenv");
dotenv.config();
module.exports = {
  faunaServerKey: process.env.FAUNA_SERVER
};

Back to the api folder create a new file called graphql.js and add the
following.

const { faunaServer } = require("./config");
const { createHttpLink } = require("apollo-link-http");
const {
  ApolloServer,
  makeRemoteExecutableSchema,
  introspectSchema
} = require("apollo-server-micro");
const fetch = require("isomorphic-fetch");

const link = createHttpLink({
  uri: "https://graphql.fauna.com/graphql",
  fetch,
  headers: {
    Authorization: `Bearer ${faunaServer}`
  }
});

const schema = makeRemoteExecutableSchema({
  schema: introspectSchema(link),
  link
});

const server = new ApolloServer({
  schema,
  playground: true,
  introspection: true
});

module.exports = (req, res, ...args) => {
  if (req.method === "OPTIONS") return res.status(200).send();

  const handler = server.createHandler();

  return handler(req, res, ...args);
};

Run now from the mySuperAPI root and deploy the API. Head back over to the app
and open the HomeScreen.js file located in the screens folder. Import gql and
the useQuery hook from Apollo. Add a query to get all the users you just created
and use the useQuery hook to display the returned data.

import { gql, useQuery } from "@apollo/client";

const GET_USERS = gql`
  query GetUsers {
    allUsers {
      data {
        name
        location
        _id
      }
    }
  }
`;

export default function HomeScreen() {

  const {
    loading,
    error,
    data
  } = useQuery(GET_USERS);

  return (
    <View style={styles.container}>
      <ScrollView
        style={styles.container}
        contentContainerStyle={styles.contentContainer}
      >
        <View style={styles.getStartedContainer}>

            {!loading && !error && data.allUsers.data.map((user) => (
              <Text key={user._id}>{user.name}</Text>
            ))}
      </ScrollView>
    </View>
  );
}

Now its time to run the app and bask in your glory!

yarn start

You have multiple options to view your app, the easiest is to just scan the QR
code on the metro bundler browser screen.

You now have a working mobile app which is pulling data with Apollo from FaunaDB
via a serverless API hosted on Vercel. Yay for technology! 🤩