Skip to content

kotify/cloud-print-utils

Repository files navigation

Cloud Print Utils

This is a collection of AWS Lambda layers and functions to render pdf documents and images from HTML.

Download layers from release section or build them yourself (requires: make, docker, zip, unzip, jq). The layers support only Amazon Linux 2023 runtimes, eg. python3.12.

Fonts

Run make build/fonts-layer.zip to build the layer. By default only dejavu fonts are installed, edit build script to install others.

Ghostscript

Ghostscript is a PDF rendering engine.

Run make build/ghostscript-layer.zip to build the layer.

WeasyPrint

WeasyPrint is python based pdf print service.

Native layer

Build the layer with:

$ make build/weasyprint-layer-python3.12.zip

To test your build:

$ make test.start.container
# a timestamp followed by "exec '/var/runtime/bootstrap'" should appear
# "docker ps" should show a running container

# in a new shell, run
$ make test.print.report
# a report.pdf file will generate to the current directory

Deploy layer:

$ aws lambda publish-layer-version \
    --region <region> \
    --layer-name <name> \
    --zip-file fileb://build/weasyprint-layer-python3.12.zip

Lambda must be configured with these env vars:

GDK_PIXBUF_MODULE_FILE="/opt/lib/loaders.cache"
FONTCONFIG_PATH="/opt/fonts"
XDG_DATA_DIRS="/opt/lib"

If you are using the release zip files ensure your Lambda instruction set architecture is set to x86_64 and not arm64.

To build a layer for python3.13 runime use:

RUNTIME=3.13 make build/weasyprint-layer-python3.13.zip

Docker Lambda

Build layer:

$ cd weasyprint
$ make build

Test layer:

$ make run

# in another terminal
$ make test
# result saved in report.pdf

Lambda Function

Simple lambda function provided, it requires BUCKET=<bucket name> env variable if files stored on s3.

Example payload to print pdf from url and return link to s3:

{"url": "https://kotify.github.io/cloud-print-utils/samples/report/", "filename": "/path/on/s3/report.pdf"}

Example paylod to print pdf from html and css data and return pdf content encoded as base64:

{"html": "<html><h1>Header</h1></html>", "css": "h1 { color: red }", "filename": "report.pdf", "return": "base64"}