Upload & fetch media files from AWS S3 in Flutter

Question

My flutter app is using firebase as a backend but I need to store media files (photos & videos) in my s3 bucket. The mission is to upload the media retrieved from the image picker into s3 & get back the url, which can then be stored as a string in my firebase database.

The problem is a scarcity of aws libraries or api for dart 2. I found 3 in pub, but 2 of them were incompatible with dart 2 & 1 was under development. Has anyone implemented this in flutter using dart 2? Any suggestions are welcome. Thank you.

The packages I found were (pub.dartlang.org) : aws_client, aws_interop, amazon_s3

Solution 1

There's a few ways to do this. One way is to sign your request with Signature V4 and POST your file to S3.

First, create a policy helper:

import 'dart:convert';
import 'package:amazon_cognito_identity_dart/sig_v4.dart';

class Policy {
  String expiration;
  String region;
  String bucket;
  String key;
  String credential;
  String datetime;
  int maxFileSize;

  Policy(this.key, this.bucket, this.datetime, this.expiration, this.credential,
      this.maxFileSize,
      {this.region = 'us-east-1'});

  factory Policy.fromS3PresignedPost(
    String key,
    String bucket,
    String accessKeyId,
    int expiryMinutes,
    int maxFileSize, {
    String region,
  }) {
    final datetime = SigV4.generateDatetime();
    final expiration = (DateTime.now())
        .add(Duration(minutes: expiryMinutes))
        .toUtc()
        .toString()
        .split(' ')
        .join('T');
    final cred =
        '$accessKeyId/${SigV4.buildCredentialScope(datetime, region, 's3')}';
    final p = Policy(key, bucket, datetime, expiration, cred, maxFileSize,
        region: region);
    return p;
  }

  String encode() {
    final bytes = utf8.encode(toString());
    return base64.encode(bytes);
  }

  @override
  String toString() {
    return '''
{ "expiration": "${this.expiration}",
  "conditions": [
    {"bucket": "${this.bucket}"},
    ["starts-with", "\$key", "${this.key}"],
    {"acl": "public-read"},
    ["content-length-range", 1, ${this.maxFileSize}],
    {"x-amz-credential": "${this.credential}"},
    {"x-amz-algorithm": "AWS4-HMAC-SHA256"},
    {"x-amz-date": "${this.datetime}" }
  ]
}
''';
  }
}

Then, sign your request with your policy helper and upload via http.MultipartRequest:

import 'dart:convert';
import 'dart:io';
import 'package:path/path.dart' as path;
import 'package:async/async.dart';
import 'package:http/http.dart' as http;
import 'package:test/test.dart';
import 'package:amazon_cognito_identity_dart/sig_v4.dart';
import './policy.dart';

void main() {
  const _accessKeyId = 'AKXXXXXXXXXXXXXXXXXX';
  const _secretKeyId = 'xxxxxxxxxxxxxxxxxxxxxxxx/xxxxxxxxxxxxxxx';
  const _region = 'ap-southeast-1';
  const _s3Endpoint =
      'https://bucketname.s3-ap-southeast-1.amazonaws.com';

  final file = File(path.join('/path/to/file', 'square-cinnamon.jpg'));
  final stream = http.ByteStream(DelegatingStream.typed(file.openRead()));
  final length = await file.length();

  final uri = Uri.parse(_s3Endpoint);
  final req = http.MultipartRequest("POST", uri);
  final multipartFile = http.MultipartFile('file', stream, length,
      filename: path.basename(file.path));

  final policy = Policy.fromS3PresignedPost('uploaded/square-cinnamon.jpg',
      'bucketname', _accessKeyId, 15, length,
      region: _region);
  final key =
      SigV4.calculateSigningKey(_secretKeyId, policy.datetime, _region, 's3');
  final signature = SigV4.calculateSignature(key, policy.encode());

  req.files.add(multipartFile);
  req.fields['key'] = policy.key;
  req.fields['acl'] = 'public-read';
  req.fields['X-Amz-Credential'] = policy.credential;
  req.fields['X-Amz-Algorithm'] = 'AWS4-HMAC-SHA256';
  req.fields['X-Amz-Date'] = policy.datetime;
  req.fields['Policy'] = policy.encode();
  req.fields['X-Amz-Signature'] = signature;

  try {
    final res = await req.send();
    await for (var value in res.stream.transform(utf8.decoder)) {
      print(value);
    }
  } catch (e) {
    print(e.toString());
  }
}

Please note that this method requires you to provide your Access Key and Secret Key. If you use a service like Cognito, it is recommended to get a temporary Access Key and Secret Key. Examples using temporary access found here.

Disclaimer: I am the original author of the Signature V4 package.

Solution 2

You can use package amazon_s3_cognito to upload and delete the images to amazon s3.

I am the author of the plugin and we are using this plugin successfully in many of our projects.

        import 'package:amazon_s3_cognito/amazon_s3_cognito.dart';
        import 'package:amazon_s3_cognito/aws_region.dart';

        String uploadedImageUrl = await AmazonS3Cognito.uploadImage(
                  _image.path, BUCKET_NAME, IDENTITY_POOL_ID);


        //Use the below code to upload an image to amazon s3 server
        //I advise using this method for image upload.
        String uploadedImageUrl = await AmazonS3Cognito.upload(
                    _image.path,
                    BUCKET_NAME,
                    IDENTITY_POOL_ID,
                    IMAGE_NAME,
                    AwsRegion.US_EAST_1,
                    AwsRegion.AP_SOUTHEAST_1)

_image.path = path of the image you want to upload (file.path method in flutter)

IMAGE_NAME = this image uploads to s3 server with the name you give here.

        //use below code to delete an image
         String result = AmazonS3Cognito.delete(
                    BUCKET_NAME,
                    IDENTITY_POOL_ID,
                    IMAGE_NAME,
                    AwsRegion.US_EAST_1,
                    AwsRegion.AP_SOUTHEAST_1)

For fetching images you could use cached_network_image package
The CachedNetworkImage can be used directly or through the ImageProvider.

    CachedNetworkImage(
            imageUrl: "http://via.placeholder.com/350x150",
            placeholder: (context, url) => new CircularProgressIndicator(),
            errorWidget: (context, url, error) => new Icon(Icons.error),
         ),`enter code here`


    Future<String> _getFilePath(Asset asset,ListingImage listingImage) async{
       try {
         if (!isUploadCancelled) {
           // getting a directory path for saving
           final directory = await getTemporaryDirectory();
           String path = directory.path;
           File file = File(path + "/temp_" + listingImage.index.toString() + "_"+DateTime.now().microsecondsSinceEpoch.toString());
           listingImage.file = file;
           file = await file.writeAsBytes( asset.imageData.buffer.asUint8List(asset.imageData.offsetInBytes, asset.imageData.lengthInBytes));

           return file.path;
         } else {
           return null;
         }
       } catch(exceptioon) {
         return null;
       }

      }