Flutter – Avoiding Jank

Parsing large JSON files can lead to poor performance of apps and shuttering animations. If parsing and computing a large JSON file take more than 16 milliseconds, the users will experience Jank. Dart by default uses a single thread to perform these tasks, though being simple and fast for less expensive computation, it fails when the computation in large.

To avoid these janks, isolates can be used to perform all computations in a different thread in the background. In this article, we will explore the process of parsing JSON in the background. To do so follow the below steps:

  1. Import the http package.
  2. Make a network request using the http package
  3. Change the response into dart objects in list form
  4. Move this work to a different isolate

Importing The http Package:

To install the http package use the below command in your command prompt:

pub get

or, if you’re using the flutter cmd use the below command:

flutter pub get



After the installation add the dependency to the pubsec.yml file as shown below:

import 'package:http/http.dart' as http;

Requesting Data:

We can use the http.get() method to fetch the sample data of 5000 images from JSONPlaceholder REST API as shown below:

Dart

filter_none

edit
close

play_arrow

link
brightness_4
code

Future<http.Response> fetchPhotos(http.Client client) async {
}

chevron_right


Parsing and Converting the Data:

We now have 5000 photos in JSON form received from the http.response and these need to be parsed and converted into a list of Dart objects. To do show First create a Photo class as shown below.

Here we will create a Photo class that contains the JSON data as shown below:

Dart

filter_none

edit
close

play_arrow

link
brightness_4
code

class Photo_list {
  final int photo_id;
  final String photo_title;
  final String photo_thumbnailUrl;
  
  Photo({this.photo_id, this.photo_title, this.photo_thumbnailUrl});
  
  factory Photo_list.fromJson(Map<String, dynamic> json) {
    return Photo_list(
      photo_id: json['id'] as int,
      photo_title: json['title'] as String,
      photo_thumbnailUrl: json['thumbnailUrl'] as String,
    );
  }
}

chevron_right


Now update the getPhoto() function to returns a Future<List<Photo>> through the below steps:

  1. Make a Photos_parser() function to convert the response body into a List<Photo>.
  2. Use the Photos_parser() function in the getPhotos() function.

Dart

filter_none

edit
close

play_arrow

link
brightness_4
code

List<Photo>Photos_parser(String responseBody) {
  final parsed = json.decode(responseBody).cast<Map<String, dynamic>>();
  
  return parsed.map<Photo>((json) => Photo.fromJson(json)).toList();
}
  
Future<List<Photo>> getPhotos(http.Client client) async {
  final response =
      await client.get('https://jsonplaceholder.typicode.com/photos');
  
  return Photos_parser(response.body);
}

chevron_right


Moving work to a different Isolate:

The compute() function can be used to move the work to a separate isolate where it will be parsed and converted in the background. It runs expensive functions in a background isolate and returns the result. 

Dart

filter_none

edit
close

play_arrow

link
brightness_4
code

Future<List<Photo>> getPhotos(http.Client client) async {
  final response =
      await client.get('https://jsonplaceholder.typicode.com/photos');
  return compute(parsePhotos, response.body);
}

chevron_right


Complete Source Code:

Dart

filter_none

edit
close

play_arrow

link
brightness_4
code

import 'dart:async';
import 'dart:convert';
  
import 'package:flutter/foundation.dart';
import 'package:flutter/material.dart';
import 'package:http/http.dart' as http;
  
Future<List<Photo>> getPhotos(http.Client client) async {
  final response =
  
  // run Photos_parser in a separate isolate.
  return compute(Photos_parser, response.body);
}
  
List<Photo> Photos_parser(String responseBody) {
  final parsed = jsonDecode(responseBody).cast<Map<String, dynamic>>();
  
  return parsed.map<Photo>((json) => Photo.fromJson(json)).toList();
}
  
class Photo_list{
  final int Photo_albumId;
  final int Photo_id;
  final String Photo_title;
  final String Photo_url;
  final String Photo_thumbnailUrl;
  
  Photo_list({this.Photo_albumId, this.Photo_id, this.Photo_title, this.Photo_url, this.Photo_thumbnailUrl});
  
  factory Photo_list.fromJson(Map<String, dynamic> json) {
    return Photo_list(
      Photo_albumId: json['albumId'] as int,
      Photo_id: json['id'] as int,
      Photo_title: json['title'] as String,
      Photo_url: json['url'] as String,
      Photo_thumbnailUrl: json['thumbnailUrl'] as String,
    );
  }
}
  
void main() => runApp(MyApp());
  
class MyApp extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    final appTitle = 'Isolate Demo';
  
    return MaterialApp(
      title: appTitle,
      home: MyHomePage(title: appTitle),
    );
  }
}
  
class HomePage extends StatelessWidget {
  final String title;
  
  HomePage({Key key, this.title}) : super(key: key);
  
  @override
  Widget build(BuildContext context) {
    return Scaffold(
      appBar: AppBar(
        title: Text(title),
      ),
      body: FutureBuilder<List<Photo>>(
        future: fetchPhotos(http.Client()),
        builder: (context, snapshot) {
          if (snapshot.hasError) print(snapshot.error);
  
          return snapshot.hasData
              ? PhotosList(photos: snapshot.data)
              : Center(child: CircularProgressIndicator());
        },
      ),
    );
  }
}
  
class PhotosList extends StatelessWidget {
  final List<Photo> photos;
  
  PhotosList({Key key, this.photos}) : super(key: key);
  
  @override
  Widget build(BuildContext context) {
    return GridView.builder(
      gridDelegate: SliverGridDelegateWithFixedCrossAxisCount(
        crossAxisCount: 2,
      ),
      itemCount: photos.length,
      itemBuilder: (context, index) {
        return Image.network(photos[index].Photo_thumbnailUrl);
      },
    );
  }
}

chevron_right


Output:




My Personal Notes arrow_drop_up


If you like GeeksforGeeks and would like to contribute, you can also write an article using contribute.geeksforgeeks.org or mail your article to contribute@geeksforgeeks.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.

Please Improve this article if you find anything incorrect by clicking on the "Improve Article" button below.


Article Tags :

Be the First to upvote.


Please write to us at contribute@geeksforgeeks.org to report any issue with the above content.