Note: Currently the deployment is down to avoid AWS charges

text-to-pdf-converter / aws-full-stack-demo-app

The text-to-PDF converter app lets users upload text files, converts them to PDF format using AWS services, and provides the PDFs for download. The app uses a serverless backend with S3 for storage, Lambda functions for processing, DynamoDB for metadata, and EC2 instances for conversion.

Application Flow

1. User Interaction:

   • User submits a text file through the frontend application.

2. Frontend Action:

   • Frontend sends a request to a proxy Lambda function to obtain a signed S3 URL for uploading the file.
   • Frontend uses the signed S3 URL to upload the user's file directly to an S3 bucket.
   • Upon successful upload, a new entry is inserted into DynamoDB, containing metadata about the uploaded file.

3. DynamoDB Trigger:

   • The DynamoDB entry triggers a Lambda function to execute.

4. Lambda Execution:

   • The Lambda function initiates the creation of an EC2 instance.

5. EC2 Instance Action:

   • The EC2 instance runs a script that converts the text file to PDF format.
   • Once converted, the PDF file is uploaded back to the same or a different S3 bucket.
   • DynamoDB entry updated with the new location and details of the converted PDF file.

6. User Interaction (Download):

   • The user can download the converted PDF file through the frontend application.

This app utilizes AWS services like S3, API Gateway, DynamoDB, Lambda, and EC2 in a structured and efficient manner.

*Note: Make sure you have Node.js installed, we will be using npm*

Backend Setup

*Note: Before deploying with AWS CDK, ensure that aws-cdk, awscli, and aws-sdk are installed. Additionally, make sure you are logged into AWS CLI using the root/IAM account you intend to deploy the CDK to*

• Navigate to the backend directory

  cd backend

• Install dependencies

  npm install

• Deploy the stack to CDK using my custom script named deploy.sh

  ./deploy.sh

  This deployment outputs the API key & API endpoints for generateSignedS3UrlApi & manageUserSubmissionsTableApi. Please note these down as you will need these for the frontend.

  For more details on how deploy.sh works and its configuration, refer to the script explanation below.

Frontend Setup

• Navigate to the frontend directory

  cd frontend

• Install dependencies

  npm install
  

• Launch the app

  npm start
  

  Be sure to populate the .env.github file with the API endpoints generated by the backend and rename it to .env.

  I've implemented a proxy Lambda function combination to securely handle API calls using the keys. If you opt not to use the proxy, remember to include your backend-generated API key as a header named x-api-key in your requests to the backend.

deploy.sh Script Explanation

Overview

The deploy.sh script automates the deployment and configuration of an AWS CDK stack named BackendStack.

Script Breakdown

Specify the AWS region for your stack deployment.

region="us-east-1"

Generate the CloudFormation templates from your AWS CDK app.

cdk synth

Sets up necessary resources (S3, Lambda, DynamoDB, etc.) for the CDK environment.

cdk bootstrap

Deploys the CloudFormation stack named BackendStack to your AWS account.

cdk deploy BackendStack

script.py Script Explanation

Overview

This python script executes within an EC2 instance triggered by events with submitter entry as "user" in DynamoDB

• Retrieves submission details from DynamoDB
• Downloads the file linked to submission from S3
• Converts the txt file to pdf and uploads back to S3
• Updates DynamoDB with updated s3 file path and submitter entry as "server"

Arguments

When the EC2 Instance is spinned off from the lambda, it is passed a user data script to execute. That script downloads this(script.py) scipt to the instance and executes it with following arguments

MANAGE_USER_SUBMISSIONS_TABLE_API => API endpoint for the manageUserSubmissionsTableApi

GENERATE_SIGNED_S3_URL_API => API endpoint for the generateSignedS3UrlApi

SUBMISSION_ID = Submission ID for which text file needs to be converted to PDF

API_KEY => API_KEY for both the API's mentioned above

Main Functions

• get_dynamodb_entry(): Retrieves submission details from DynamoDB and downloads files from S3

• get_signed_s3_url(): Gets signed URL from generateSignedS3UrlApi for uploading/downloading files using the API_KEY

• download_file_from_s3(): Downloads file from S3 bucket using the signed URL

• upload_file_to_s3(): Uploads file to S3 bucket using the signed URL

• convert_to_pdf(): Converts the downloaded text file to pdf

• insert_file_to_dynamodb(): Updates the S3 path for the pdf and submitter entry as "server" for the current submission

• process_submission(): Handles above functions in order

Note

The submitter attribute with a value of "user" in DynamoDB ensures that the EC2 instance is launched only when a user submits a text file, not when the script itself performs operations. Conversely, when the script submits data, it marks the entry with "server" as the submitter. This differentiation clarifies that the EC2 instance is triggered specifically in response to user submissions, distinguishing it from automated script actions and avoiding an infinite loop.

API's

generateSignedS3UrlApi

The generateSignedS3UrlApi is attached to a lambda function named generateSignedS3Url which generates signed URLs for secure access to objects stored in an Amazon S3 bucket. These URLs grant temporary, controlled access to specified resources without exposing AWS credentials.

manageUserSubmissionsTableApi

The manageUserSubmissionsTableApi is attached to a lambda function named manageUserSubmissionsTable which is designed to manage submissions within a database table named userSubmissionsTable. This API allows operations such as creating new submissions, retrieving submission details, updating existing submissions.