Fast, Enjoyable & Secure NoSQL Database

Hive is a lightweight and blazing fast key-value database written in pure Dart. Inspired by Bitcask.

Tutorial

For a quick tutorial, see this wonderful post by Tijn van den Eijnde

Features

• 🚀 Cross platform: mobile, desktop, browser
• ⚡ Great performance (see benchmark)
• ❤️ Simple, powerful, & intuitive API
• 🔒 Strong encryption built in
• 🎈 NO native dependencies
• 🔋 Batteries included

New features in Hive CE

Hive CE is a spiritual continuation of Hive v2 with the following new features:

• Flutter web WASM support
• Support for Sets
• A built in Duration adapter
• A HiveRegistrar extension that lets you register all your generated adapters in one call
• Support for constructor parameter defaults
• Freezed support
• Support for generating adapters with classes that use named imports
• Automatic type adapter generation using the GenerateAdapters annotation
  • No more manually adding annotations to every type and field
  • Generate adapters for classes outside the current package

Hive CE (v2) vs Hive v4 (Isar)

You may be considering attempting to make the dev version of Hive v4 work in your project. I strongly advise against this. Not only is Hive v4 not stable, but it is also much slower and less efficient than Hive CE.

This is a comparison of the time to complete a given number of write operations and the resulting database file size:

Operations	Hive CE		Hive v4
10	0.00 s	0.00 MB	0.01 s	1.00 MB
100	0.01 s	0.01 MB	0.01 s	1.00 MB
1,000	0.05 s	0.11 MB	0.07 s	1.00 MB
10,000	0.21 s	1.10 MB	0.69 s	5.00 MB
100,000	1.53 s	10.97 MB	7.01 s	30.00 MB
1,000,000	19.97 s	109.67 MB	89.81 s	290.00 MB

Database size in Hive v4 is directly affected by the length of field names in model classes which is not ideal. Also Hive v4 is much slower than Hive CE for large numbers of operations.

The benchmark was performed on an M3 Max MacBook Pro. You can see the benchmark code here.

Migrating from Hive v2

The hive_ce package is a drop in replacement for Hive v2. Make the following replacements in your project:

pubspec.yaml

# old dependencies
dependencies:
  hive: ^2.0.0
  hive_flutter: ^1.0.0

dev_dependencies:
  hive_generator: ^1.0.0

# new dependencies
dependencies:
  hive_ce: latest
  hive_ce_flutter: latest

dev_dependencies:
  hive_ce_generator: latest

Dart files

// old imports
import 'package:hive/hive.dart';
import 'package:hive_flutter/hive_flutter.dart';

// new imports
import 'package:hive_ce/hive.dart';
import 'package:hive_ce_flutter/hive_flutter.dart';

Usage

You can use Hive just like a map. It is not necessary to await Futures.

import 'package:hive_ce/hive.dart';

void example() {
  final box = Hive.box('myBox');
  box.put('name', 'David');
  final name = box.get('name');
  print('Name: $name');
}

BoxCollections

BoxCollections are a set of boxes which can be similarly used as normal boxes, except of that they dramatically improve speed on web. They support opening and closing all boxes of a collection at once and more efficiently store data in indexed DB on web.

Aside, they also expose Transactions which can be used to speed up tremendous numbers of database transactions on web.

On dart:io platforms, there is no performance gain by BoxCollections or Transactions. Only BoxCollections might be useful for some box hierarchy and development experience.

import 'package:hive_ce/hive.dart';
import 'hive_cipher_impl.dart';

void example() async {
  // Create a box collection
  final collection = await BoxCollection.open(
    // Name of your database
    'MyFirstFluffyBox',
    // Names of your boxes
    {'cats', 'dogs'},
    // Path where to store your boxes (Only used in Flutter / Dart IO)
    path: './',
    // Key to encrypt your boxes (Only used in Flutter / Dart IO)
    key: HiveCipherImpl(),
  );

  // Open your boxes. Optional: Give it a type.
  final catsBox = await collection.openBox<Map>('cats');

  // Put something in
  await catsBox.put('fluffy', {'name': 'Fluffy', 'age': 4});
  await catsBox.put('loki', {'name': 'Loki', 'age': 2});

  // Get values of type (immutable) Map?
  final loki = await catsBox.get('loki');
  print('Loki is ${loki?['age']} years old.');

  // Returns a List of values
  final cats = await catsBox.getAll(['loki', 'fluffy']);
  print(cats);

  // Returns a List<String> of all keys
  final allCatKeys = await catsBox.getAllKeys();
  print(allCatKeys);

  // Returns a Map<String, Map> with all keys and entries
  final catMap = await catsBox.getAllValues();
  print(catMap);

  // delete one or more entries
  await catsBox.delete('loki');
  await catsBox.deleteAll(['loki', 'fluffy']);

  // ...or clear the whole box at once
  await catsBox.clear();

  // Speed up write actions with transactions
  await collection.transaction(
    () async {
      await catsBox.put('fluffy', {'name': 'Fluffy', 'age': 4});
      await catsBox.put('loki', {'name': 'Loki', 'age': 2});
      // ...
    },
    boxNames: ['cats'], // By default all boxes become blocked.
    readOnly: false,
  );
}

Store objects

Hive not only supports primitives, lists, and maps but also any Dart object you like. You need to generate type adapters before you can store custom objects.

Create model classes

import 'package:hive_ce/hive.dart';

class Person extends HiveObject {
  Person({required this.name, required this.age});

  String name;
  int age;
}

Create a GenerateAdapters annotation

Usually this is placed in lib/hive/hive_adapters.dart

import 'package:hive_ce/hive.dart';
import 'person.dart';

part 'hive_adapters.g.dart';

@GenerateAdapters([AdapterSpec<Person>()])
// Annotations must be on some element
// ignore: unused_element
void _() {}

Update pubspec.yaml

dev_dependencies:
  build_runner: latest
  hive_ce_generator: latest

Run build_runner

dart pub run build_runner build --delete-conflicting-outputs

This will generate the following:

• TypeAdapters for the specified AdapterSpecs
• TypeAdapters for all explicitly defined HiveTypes
• A hive_adapters.g.dart file containing all adapters generated from the GenerateAdapters annotation
• A hive_adapters.g.yaml file
• A hive_registrar.g.dart file containing an extension method to register all generated adapters

All of the generated files should be checked into version control. These files are explained in more detail below.

Use the Hive registrar

The Hive Registrar allows you to register all generated TypeAdapters in one call

import 'dart:io';
import 'package:hive_ce/hive.dart';
import 'package:your_package/hive/hive_registrar.g.dart';

void main() {
  Hive
    ..init(Directory.current.path)
    ..registerAdapters();
}

Using HiveObject methods

Extending HiveObject is optional but it provides handy methods like save() and delete().

import 'package:hive_ce/hive.dart';
import 'person.dart';

void example() async {
  final box = await Hive.openBox('myBox');

  final person = Person(name: 'Dave', age: 22);
  await box.add(person);

  print(box.getAt(0)); // Dave - 22

  person.age = 30;
  await person.save();

  print(box.getAt(0)); // Dave - 30
}

About hive_adapters.g.yaml

The Hive schema is a generated yaml file that contains the information necessary to incrementally update the generated TypeAdapters as your model classes evolve.

Some migrations may require manual modifications to the Hive schema file. One example is class/field renaming. Without manual intervention, the generator will see both an added and removed class/field. To resolve this, manually rename the class/field in the schema.

Migrating to GenerateAdapters

If you already have model classes with HiveType and HiveField annotations, you can take the following steps to migrate to the new GenerateAdapters annotation:

1. Convert all default values to constructor parameter defaults
2. Add the following to your build.yaml file:

targets:
  $default:
    builders:
      hive_ce_generator|hive_schema_migrator:
        enabled: true

3. Run the build_runner. This will generate lib/hive/hive_adapters.dart and lib/hive/hive_adapters.g.yaml.
4. Revert the build.yaml changes
5. Remove all explicit HiveType and HiveField annotations from your model classes
6. Run the build_runner again

Explicitly defining HiveTypes

The old method of defining HiveTypes is still supported, but should be unnecessary now that Hive CE supports constructor parameter defaults. If you have a use-case that GenerateAdapters does not support, please create an issue on GitHub.

Unfortunately it is not possible for GenerateAdapters to handle private fields. You can use @protected instead if necessary.

Add fields to objects

When adding a new non-nullable field to an existing object, you need to specify a default value to ensure compatibility with existing data.

For example, consider an existing database with a Person object:

import 'package:hive_ce/hive.dart';

class Person extends HiveObject {
  Person({required this.name, required this.age});

  String name;
  int age;
}

If you want to add a balance field, you must specify a default value or else reading existing data will result in null errors:

import 'package:hive_ce/hive.dart';

class Person extends HiveObject {
  Person({required this.name, required this.age, this.balance = 0});

  String name;
  int age;
  double balance;
}

After modifying the model, remember to run build_runner to regenerate the TypeAdapters

Hive ❤️ Flutter

Hive was written with Flutter in mind. It is a perfect fit if you need a lightweight datastore for your app. After adding the required dependencies and initializing Hive, you can use Hive in your project:

import 'package:hive_ce/hive.dart';
import 'package:hive_ce_flutter/hive_flutter.dart';

class SettingsPage extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    return ValueListenableBuilder(
      valueListenable: Hive.box('settings').listenable(),
      builder: (context, box, widget) {
        return Switch(
          value: box.get('darkMode'),
          onChanged: (val) {
            box.put('darkMode', val);
          }
        );
      },
    );
  }
}

Boxes are cached and therefore fast enough to be used directly in the build() method of Flutter widgets.

Benchmark

1000 read iterations	1000 write iterations
SharedPreferences is on par with Hive when it comes to read performance. SQLite performs much worse.	Hive greatly outperforms SQLite and SharedPreferences when it comes to writing or deleting.

The benchmark was performed on a Oneplus 6T with Android Q. You can run the benchmark yourself.

*Take this benchmark with a grain of salt. It is very hard to compare databases objectively since they were made for different purposes.

Licence

Copyright 2019 Simon Leier

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

   http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.