Serverless LibreOffice

AWS Lambda DOCX to PDF converter using LibreOffice

This project was created based on https://github.com/shelfio/aws-lambda-libreoffice which should be referenced for documentation, code samples, support, or other issues.

Installation

npm install

Development

Make changes to handler.js as needed

Deployment

For a first time deployment, you will need to set up an AWS ECR repository for the Docker image:

aws ecr get-login-password --region <AWS REGION> | docker login --username AWS --password-stdin <AWS ACCOUNT ID>.dkr.ecr.ca-central-1.amazonaws.com

aws ecr create-repository --repository-name docx-to-pdf-converter --image-scanning-configuration scanOnPush=true --image-tag-mutability MUTABLE

Once the ECR repository has been created, run the following to build, tag, and push the image to ECR

docker build -t docx-to-pdf-converter .

docker tag  docx-to-pdf-converter:latest <AWS ACCOUNT ID>.dkr.ecr.ca-central-1.amazonaws.com/docx-to-pdf-converter:latest

docker push <AWS ACCOUNT ID>.dkr.ecr.ca-central-1.amazonaws.com/docx-to-pdf-converter:latest

[First time deployment only] Once the image has been pushed, you'll need to create a Lambda function from the image. To do this go to Lambda > Create Function > Container Image > Browse images and select your newly created image.

[Subsequent deployments] Repeat the build, tag, and push process, and then go to your Lambda's Image settings. Select Deploy new image and browse for the image tagged latest

Usage

The handler function accepts a bucket and a key argument, which are the name of an S3 bucket, and the S3 key of an object in that bucket. key should point to a valid .docx document.

import AWS from "aws-sdk";

async function convertToPdf(): Promise<string> {
  const lambda = new AWS.Lambda();
  const payload = {
      bucket: 'bucket_name',
      key: 's3_key.docx',
  };
  const params: AWS.Lambda.InvocationRequest = {
    FunctionName: 'docx-to-pdf-converter',
    Payload: JSON.stringify(payload),
    InvocationType: 'RequestResponse',
  };
  const response = await lambda
    .invoke(params, (err) => {
      if (err) {
        throw err;
      }
    })
    .promise();

  const result = JSON.parse(response.Payload as string);
  if (!('Key' in result)) {
    throw new Error('File could not be converted to PDF');
  }
  return result.Key;
}

Troubleshooting

• If you receive an AccessDenied error when reading/writing to S3, check that the Lambda function has permission to read/write to the bucket and object that you supplied as input.