s3-cognito-cli

This is a simple reference architecture and demo/example to enable external users managed by Cognito User pool to access an internal S3 bucket.

The upload logic uses S3 multiUpload feature.

index.js => entry point
configer.js => set the local configuration
uploader.js => s3 upload logic
saveUploadJob method in uploader => an additional example to save the upload object's meta data through AWS AppSync. Check 3 about the option.

1. Creating a S3 Bucket and Cognito User Pool with amplify cli

Generate configuration and resources

Option 1: Use AWS AMPLIFY CLI to create resources
Option 2: Use AWS Management Console to create resources

2. Quick Start

2-1. Installation - only once

$ npm install -g s3-cognito-cli

2-2. Profile Configuration - which S3 Bucket, which Cognito Userpool to use - only once for one S3 & Cognito environment. <aws-exports.json> is created in Step 1.

One time job, you only need to save a config once.

$ s3cognito config --set <aws-exports.json> --usr <user> --pwd <pwd>

2-3. SignUp a Cognito Userpool User - only once if only using this user

One time job, you only need to sign up once for a certain user.

$ s3cognito signup --usr <user> --pwd <pass> --email <email_address>

2-4. Confirm SignUp - only once for this user

One time job, you only need to confirm once.

$ s3cognito confirm --usr <user> --code <confirmation_code>

2-5. Upload your file

After the previous steps you are able to upload different files as many times as you want with this upload command. Under the hood the program will first sign in through Cognito User Pool with the Amplify library and then multi-upload the local file to the prepared S3 bucket.

$ s3cognito upload <your_file_relative_path>

3. Save upload task and object meta data(Optional)

To save the original file name and the owner of the upload task it is recommended to keep a record. This saveUploadJob method uses AppSync as a facade of a DynamoDB table.

const saveUploadJob = async ({objKey, fileKey, fileName, user}) => {
	const API = Amplify.API;
	const graphqlOperation = Amplify.graphqlOperation;

	const input = {
	    user,
	    objKey,
	    fileKey,
	    fileName,
	    status: 'Done',
	    uploadedOn: new Date().toISOString()
	};

    const rtn = await API.graphql(graphqlOperation(createUploadJob, { input }))
      .catch(e => {
        throw e;
      });

  	return { fileKey, data: rtn };
}

It will create a record in DynamoDB with the following information:

user: user name of the upload job's owner
objKey: the key of the uploaded file as an object on s3 including full path
fileKey: the s3 object's key without the full path
fileName: original file name
status: upload status
uploadedOn: job's timestamp

const createUploadJob = `mutation CreateUploadJob($input: CreateUploadJob!) {
  createUploadJob(input: $input) {
    user
    objKey
    fileKey
    fileName
    status
    uploadedOn
  }
}
`;

This repository does not include the steps to create the AppSync API and the DynamoDB table. Please find this following official tutorial as reference. AppSync Tutorial: DynamoDB Resolvers

komushi/s3-cognito-cli