Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/barchart/aws-lambda-pdf-generator/llms.txt

Use this file to discover all available pages before exploring further.

The PDF Generator can be exercised at two distinct stages of the development lifecycle: locally before any AWS resources exist, and against the live deployed endpoint after a successful yarn deploy:prod or yarn deploy:stage. Local invocation uses the Serverless CLI to simulate the Lambda execution environment on your machine, which is useful for rapid iteration. Post-deployment testing validates the full AWS path — API Gateway, Lambda cold start, Chromium layer, and S3 large-response handling — using plain HTTP clients such as curl.

​
Local Testing

Local invocation runs the Lambda handler directly on your machine using the Serverless framework’s invoke local command. No AWS credentials or live infrastructure are needed; the framework resolves environment variables from serverless.yml and loads the handler from source.
1

Navigate to the API package

All sls commands must be run from within the packages/api directory, where serverless.yml is located:
cd ./packages/api
2

Invoke with the standard test payload

Run the print function using the bundled test event file:
sls invoke local -f print --path print/test/print.json
The file print/test/print.json contains a pre-built API Gateway proxy event. Its body field is a base64-encoded JSON string that decodes to:
{
  "source": "testing",
  "html": "<html><body><h1>Hello World</h1><div>You have converted an HTML document into a PDF file.</div></body></html>"
}
The isBase64Encoded: true flag instructs the handler to decode the body before passing it to Puppeteer, mirroring exactly how API Gateway forwards binary-encoded POST requests.
3

Invoke with the 5 MB stress-test payload

To verify that large-payload handling works correctly (including the S3 fallback path for responses that exceed Lambda’s 6 MB synchronous response limit), use the 5 MB test fixture:
sls invoke local -f print --path print/test/print.5MB.json
This payload contains a significantly larger HTML document and exercises the code path that writes the PDF to S3 and returns a pre-signed URL instead of embedding the binary directly in the Lambda response.

​
Testing the Deployed Endpoint

After a successful deployment, the Serverless framework prints the base URL of your API Gateway stage to the console. Use that URL to send live HTTP requests.

​
Convert HTML to PDF — POST /print

Send an HTML document in the request body and save the binary PDF response to a local file: 
curl 'https://{api-identifier}.execute-api.us-east-1.amazonaws.com/{stage}/print' \
  -X 'POST' \
  --output result.pdf \
  --header 'Content-Type: application/json' \
  --data-binary '{"source":"testing","html":"<html><body><h1>Hello World</h1><div>You have converted an HTML document into a PDF file.</div></body></html>"}'
The --output result.pdf flag writes the raw PDF bytes returned by the service directly to disk. Open result.pdf in any PDF viewer to confirm a successful conversion.

​
Read service metadata — GET /service/version

Verify that the service is reachable and inspect its version metadata: 
curl 'https://{api-identifier}.execute-api.us-east-1.amazonaws.com/{stage}/service/version'
A successful response returns a JSON object containing the deployed service version.

​
Interpreting Results

HTTP StatusMeaningCommon cause
200 OKPDF generated successfullyThe response body is a binary application/pdf document
400 Bad RequestRequest validation failedThe required html field is missing or the request body is malformed JSON
500 Internal Server ErrorServer-side processing errorChromium failed to launch, Puppeteer encountered a rendering error, or the S3 large-response bucket is unreachable
Replace {api-identifier} and {stage} in the example URLs with the actual values printed by the Serverless framework after deployment. For example, a production endpoint might look like https://abc123xyz.execute-api.us-east-1.amazonaws.com/prod/print.
If a request returns a 500 error or an unexpected response, check the Lambda function’s CloudWatch Log Group for the full stack trace. Logs are retained for 90 days and are located at:/aws/lambda/serverless-pdf-generator-{stage}-printYou can tail the logs in real time using the AWS CLI:
aws logs tail /aws/lambda/serverless-pdf-generator-{stage}-print --follow

Build docs developers (and LLMs) love

Get started for freeTalk to us