Search packages, repositories and app categories

Package detailsv1.2.0

functional_data

Annotations and helpers for lenses for code generated with the `function_data_generator` package.

AndroidiOSWebmacOSWindowsLinux
30-day downloads
181K
Likes
19
Pub points
150

Install package

flutter pub add functional_data

Package Overview

Package selection guidance compiled from public package information, community data, and project usage.

FunctionalData is a Dart package that provides annotations and utility helpers for creating lenses—functional programming constructs used to access and modify nested data structures. It works in conjunction with the `function_data_generator` code generator to automatically produce boilerplate-free, immutable data classes with powerful update capabilities. The package enables safe, composable updates to deeply nested objects without mutation, promoting functional programming principles in Flutter and Dart applications. Ideal for state management and data transformation workflows.

Use cases

  • Immutable state updates
  • Nested data manipulation
  • State management in Flutter apps
  • Functional programming patterns
  • Code generation for data classes

Key features

  • Automatic lens generation
  • Immutability support
  • Deep nested updates
  • Type-safe transformations
  • Integration with code generators

Suitable for

  • Developers using functional programming
  • Flutter apps with complex state
  • Teams prioritizing immutability
  • Code generation workflows
  • Data-heavy applications

Considerations

  • Requires `function_data_generator`
  • Learning curve for lenses
  • Limited runtime reflection
  • Not suitable for simple data models
  • Additional build step needed

Original README

Original README snapshot from the linked source repository. Visit the repository for the latest content.

Functional Data

Build

Simple and non-intrusive code generator for boilerplate of data types. The package generates a simple mixin with operator==, hashCode, copyWith, toString, as well as lenses.

Boiler plate

Because the boiler plate is generated as a mixin, it is minimally intrusive on the interface of the class. You only have to provide a constructor with named arguments for all fields and extend the generated mixin.

@FunctionalData()
class Person extends $Person {
  final String name;
  final int age;

  const Person({this.name, this.age});

  const Person.anonymous() : this(name: "John Doe", age: null);

  int get ageInDays => numberOfDaysInMostYears * age;

  static const numberOfDaysInMostYears = 356;
}

Because of this design, you have complete control over the class. You can, for example, add named constructors or methods to the class like you're used to.

Lenses

For every class, lenses are generated for all fields which allow viewing a field or creating a new instance of the classes with that field modified in some way. For example, the lens of Person's name is Person$.name. To focus a lens on a specific instance use its of method:

final teacher = Person(name: "Arthur", age: 53);

print(Person$.name.of(teacher).value);
// -> "Arthur"

print(Person$.age.of(teacher).update(60);
// -> Person(name: "Arthur", age: 60)

print(Person$.name.of(teacher).map((name) => name.toUpperCase()));
// -> Person(name: "ARTHUR", age: 53)

This isn't very exciting yet. The power of lenses comes to light when you combine them. It allows you to easily create a copy of a large nested data structure with one of the fields in a leaf modified. Two lenses can be chained using then.

class Course extends $Course {
  final String name;
  final List<Person> students;

  const Course({this.students});
}

final programming = Course(name: "Programming 101", students: [Person(name: "Jane", age: 21), Person(name: "Tom", age: 20)]);

final firstStudentsName = Course$.students.then(List$.first<Person>()).then(Person$.name);

print(firstStudentsName.of(programming).update("Marcy"));
// -> Course(students: [Person(name: "Marcy", age: 21), Person(name: "Tom", age: 20)]

Compare this with the alternative:

final firstStudent = programming.students.first;
final updatedFirstStudent = Person(name: "Marcy", age: firstStudent.age);
final updatedStudents = [updatedFirstStudent] + programming.students.skip(1);
final updatedCourse = Course(name: programming.name, students: updatedStudents);

This is much less readable and error prone. Imagine what happens when one of the classes gains a field.

Full example:
// lens.dart
import 'package:collection/collection.dart';

import 'package:functional_data/functional_data.dart';

part 'lens.g.dart';

// Only requirement is that it has a constructor with named arguments for all fields
@FunctionalData()
class Foo extends $Foo {
  final int number;
  final String name;

  const Foo({this.number, this.name});
}

@FunctionalData()
class Bar extends $Bar {
  final Foo foo;

  @CustomEquality(DeepCollectionEquality())
  final List<Foo> foos;

  final String driver;

  const Bar({this.foo, this.foos, this.driver});
}

void main() {
  final foo = Foo(number: 42, name: "Marvin");
  final bar = Bar(foo: foo, foos: [Foo(number: 101, name: "One"), Foo(number: 102, name: "Two")], driver: "One");

  final fooName = Bar$.foo.then(Foo$.name);
  // print(fooName.map((name) => name.toUpperCase(), bar));
  print(fooName.of(bar).map((name) => name.toUpperCase()));
  // Bar(foo: Foo(number: 42, name: MARVIN), foos: [Foo(number: 101, name: One), Foo(number: 102, name: Two)], driver: One)

  final firstFooName = Bar$.foos.then(List$.atIndex<Foo>(0)).then(Foo$.name);
  // print(firstFooName.update(bar, "Twee"));
  print(firstFooName.of(bar).update("Twee"));
  // Bar(foo: Foo(number: 42, name: Marvin), foos: [Foo(number: 101, name: Twee), Foo(number: 102, name: Two)], driver: One)

  final nameOfFooNamedTwo = Bar$.foos.then(List$.where<Foo>((foo) => foo.name == "Two")).then(Foo$.name);
  print(nameOfFooNamedTwo.update(bar, "Due"));
  // Bar(foo: Foo(number: 42, name: Marvin), foos: [Foo(number: 101, name: One), Foo(number: 102, name: Due)], driver: One)

  final driversNumber =
      Bar$.foos.thenWithContext((bar) => List$.where<Foo>((foo) => foo.name == bar.driver).then(Foo$.number));
  print(driversNumber.of(bar).value);
  // 101
}

Code generation configuration

Class level configuration

To specify which features should be generated for the class, you can send arguments to @FunctionalData generator.

Example:

@FunctionalData(
  generateCopy: false,
  generateLenses: false,
)
class Foo extends _$Foo {}
Project level configuration

To specify which features should be generated for you whole project, create a file called functional_data_options.yaml in the root of your project. Class specific arguments will override the project level configuration.

Example with all possible configurations:

generateCopyWith: false
generateCopyUsing: false
generateLenses: false

Contributing

Setup

This project uses melos to manage all the packages inside this repo.

  1. Install melos via the top level workspace dart pub get.
  2. Setup melos to point to the dependencies in your local folder: melos bootstrap

Testing the Code generator:

  • Either run melos generate:dart in the top level folder, or dart run build_runner build in the example folder.

Categories

Dependencies

Dependency list 2 items

Source repository

Projects using this package