lambda-packer

lambda-packer

A streamlined tool for managing and packaging Python AWS Lambda functions

Overview

lambda-packer is a command-line tool designed to simplify the process of packaging Python AWS Lambda functions. It provides an opinionated approach to develop Lambdas using a monorepo, allowing packaging as either zip files or Docker containers, with shared dependencies packaged as Lambda layers.

Key Features

• Package Lambdas as zip files or Docker containers
• Support for multiple Lambda layers shared across functions
• Simple YAML configuration to manage Lambdas and layers
• Layer packaging with automatic dependency handling

Installation

pip install lambda-packer

Usage

1. Initialize a new repo

The init command creates a basic repository structure for your Lambda functions, including a common folder for shared dependencies, an example Lambda function, and a package_config.yaml file.

lambda-packer init <parent_directory> --lambda-name <lambda_name>

Example:

lambda-packer init my_project --lambda-name my_lambda

This command creates:

my_project/
├── common/
├── my_lambda/
│   ├── lambda.py
│   └── requirements.txt
├── dist/
└── package_config.yaml

2. Configuration

The package_config.yaml file is where you define how to package your Lambdas. You can specify the type of packaging (zip or docker), the Python runtime, and any layers associated with the Lambda.

Example package_config.yaml

lambdas:
  my_lambda:
    type:
      - zip
    file_name: lambda
    function_name: lambda_handler
    runtime: '3.12'
    platforms: ['linux/arm64', 'linux/x86_64']
    layers:
      - common

3. Package Lambda as a Zip

To package a Lambda function (for a zip type Lambda), use the following command:

lambda-packer package my_lambda

This will package the Lambda function and any referenced layers (e.g., common) into a zip file in the dist directory.

4. Package Lambda as a Docker Container

To package a Lambda as a Docker container (for a docker type Lambda), modify the package_config.yaml and set type: docker.

lambdas:
  my_lambda:
    type: docker
    runtime: "3.9"
    layers:
    - common

Then run:

lambda-packer package my_lambda

Or package them all:

layer-packer package

The tool will build a Docker image using the specified Python runtime and package the Lambda function.

5. Packaging Lambda Layers

If you need to package shared dependencies (like the common folder) as Lambda layers, you can use the package-layer command:

lambda-packer package-layer common

This command packages the common directory as a Lambda layer and zips it to the dist/ folder.

Available Commands

• init <parent_directory> --lambda-name <lambda_name>: Initialize a new monorepo with a common folder, a lambda, and package_config.yaml.
• package <lambda_name>: Package the specified Lambda function (either as zip or Docker container).
• package-layer <layer_name>: Package a specific layer (e.g., common) into a zip file.
• config <lambda_name>: Generate a package_config.yaml from an existing monorepo.
• clean: Clean the dist/ directory by removing all contents.

Example Workflow

• Initialize the project:

lambda-packer init my_project --lambda-name my_lambda

• Edit package_config.yaml to configure the Lambda:

lambdas:
  my_lambda:
    type: zip
    runtime: "3.9"
    layers:
    - common

• Install dependencies for my_lambda by editing my_lambda/requirements.txt.

• Package the Lambda:

lambda-packer package my_lambda

• Package the common layer (if needed):

lambda-packer package-layer common

6. Adding a new lambda to an existing repository

You can add a new Lambda to an existing repository using the lambda command. You can also specify layers to be added to the new Lambda.

lambda-packer lambda <lambda_name> --runtime <runtime_version> --type <zip|docker> --layers <layer1> --layers <layer2>

Example:

lambda-packer lambda my_new_lambda --runtime 3.9 --type docker --layers common --layers shared

This will create a new Lambda directory and update the package_config.yaml like so:

lambdas:
  my_new_lambda:
    runtime: "3.9"
    type: docker
    layers:
    - common
    - shared

If no layers are specified, the layers key will not be added.

Example without layers:

lambda-packer lambda my_new_lambda --runtime 3.9 --type docker

This will update the package_config.yaml like this:

lambdas:
  my_new_lambda:
    runtime: "3.9"
    type: docker

Contributing

Contributions are welcome! If you'd like to contribute to this project, please open a pull request or issue on GitHub.

Development Setup

Clone this repository and run:

git clone https://github.com/calvernaz/lambda-packer.git
cd lambda-packer
pip install -e .

For development:

pip install -e .[dev]

Running Tests

pytest tests/

Release

Bump patch version:

bumpversion patch

Push tags:

git push origin main --tags

License

This project is licensed under the MIT License.

Contact

For any questions or feedback, feel free to open an issue on GitHub.