In this quickstart, you will create and deploy a small sample database and access it from a Flutter frontend.
Prerequisites
To complete this quickstart, you'll need the following:
- An environment with the following tools installed:
- Node.js with a package manager of your choice. This tutorial assumes
npm. - The Firebase CLI
:
npm i -g firebase-tools@latest - The Flutter command line tool .
- The FlutterFire CLI
dart pub global activate flutterfire_cli
- Node.js with a package manager of your choice. This tutorial assumes
- A Google Account.
Tutorial
1. Create a new Firebase project
Start by creating a new Firebase project in the Firebase console . Then, upgrade the project to the Blaze plan.
2. Initialize your project
Create a new directory and initialize a Firebase project in it. When prompted, choose the following options:
- Choose the project you created in the previous step.
- Don't create a schema with Gemini (in this tutorial, you'll use a pre-built example schema).
- Create a new Cloud SQL instance.
- Create a Flutter template.
mkdir myproj ; cd myprojfirebase login --reauthfirebase init dataconnect
3. Review the example GraphQL definitions
In Data Connect , you define all of your database schemas and operations using GraphQL. When you initialized your project, the Firebase CLI created some example definitions to get you started.
type Movie @ table { title : String ! imageUrl : String ! genre : String } type MovieMetadata @ table { movie : Movie ! @ unique rating : Float releaseYear : Int description : String }
query ListMovies @auth ( level : PUBLIC ) { movies { id title imageUrl genre } }
4. Deploy your schemas and operations
Whenever you make changes to your database schemas, queries, or mutations, you must deploy them for your changes to take effect on the database.
firebase deploy --only dataconnect
5. Seed the database with sample data
This seed data will give you something to look at when you test the sample app. Note that in this step you are executing arbitrary GraphQL, which is allowed for administrative tasks.
firebase dataconnect:execute dataconnect/seed_data.gql
6. Generate a Dart client SDK
This command uses your GraphQL definitions to generate a Dart library specifically for your database. You use this library in your client app to perform all database operations.
You can generate libraries for multiple platforms, including Kotlin for Android,
Swift for iOS, and JavaScript for web, by adding definitions to connector.yaml
.
firebase dataconnect:sdk:generate
class ExampleConnector { ListMoviesVariablesBuilder listMovies () { return ListMoviesVariablesBuilder ( dataConnect ); } // ... }
7. Set up a Flutter app
Run these commands to set up the Flutter app to use your Firebase project.
When prompted by the flutterfire
command, select the Firebase project
you created earlier, and choose the platforms you want to target.
cd flutter_appflutter pub add firebase_core firebase_data_connectflutterfire configure
8. Write a sample Flutter client
Replace the contents of flutter_app/lib/main.dart
with this simple
Flutter app.
Notice that the app completes the necessary database access using a function from the generated SDK.
import 'package:firebase_core/firebase_core.dart' ; import 'package:firebase_data_connect/firebase_data_connect.dart' ; import 'package:flutter/material.dart' ; import 'package:flutter_app/dataconnect_generated/generated.dart' ; import 'package:flutter_app/firebase_options.dart' ; class MyApp extends StatelessWidget { late final Future<QueryResult<ListMoviesData , void >> _movieListFuture ; MyApp ({ super . key }) { _movieListFuture = ExampleConnector . instance . listMovies () . execute (); } @override Widget build ( BuildContext context ) { return MaterialApp ( home: FutureBuilder ( future: _movieListFuture , builder: ( context , snapshot ) { if ( snapshot . connectionState == ConnectionState . done ) { return ListView . builder ( scrollDirection: Axis . vertical , itemBuilder: ( context , index ) = > Card ( child: Text ( snapshot . data ? . data . movies [ index ]. title ?? "" , ), ), itemCount: snapshot . data ? . data . movies . length ?? 0 , ); } return const CircularProgressIndicator (); }, ), ); } } Future<void> main () async { await Firebase . initializeApp ( options: DefaultFirebaseOptions . currentPlatform , ); runApp ( MyApp ()); }
9. Try the app
Start the development server to see the example app in action.
flutter run
Next steps
Try the Visual Studio Code extension
When developing with Data Connect , we strongly recommend using the Visual Studio Code extension . Even if you don't use Visual Studio Code as your primary development environment, the extension provides several features that make schema and operation development more convenient:
- A GraphQL language server, providing syntax checking and autocomplete suggestions specific to Data Connect
- CodeLens buttons in line with your code that let you read and write data from your schema definition files and execute queries and mutations from your operation definitions.
- Automatically keep your generated SDKs synchronized with your GraphQL definitions.
- Simplified local emulator setup.
- Simplified deployment to production.
Use the Data Connect emulator for local development
Although this tutorial showed you how to deploy Data Connect schemas and operations directly to production, you will likely not want to make changes to your production database while you are actively developing your app. Instead, set up the Data Connect emulator and do your development work against it rather than production. The emulator sets up a local PGlite instance that behaves similarly to a live Postgres instance on CloudSQL.
Learn how to write schemas and operations for your app
When developing apps with Data Connect , the design of your schemas and operations is one of the first and most important development tasks you will complete.
- Gemini in the Firebase console is an AI tool that can generate Data Connect schemas from a natural language description of your app. This tool can get you started very quickly, especially if you've never worked with relational databases before.
- Alternatively, you can write database schemas, queries, and mutations directly using GraphQL. Start with the Design Data Connect schemas page and continue to the follow-up pages to learn how to write operations.
