poetry-aws-sam 1.4.3

Poetry plugin to work with AWS's SAM serverless tool

Description

CLI tool to run the sam build on a poetry-based project.

AWS's SAM made use of a requirements.txt file for third party dependency management. This method is limiting in comparison to other tools such as Poetry that manage sub-dependencies' versions through a lock file. This poetry plugin deploys a serverless app through SAM with dependencies being managed through Poetry.

Created by Pinnacle Solutions Group, www.pinnsg.com.

Set up

This plugin is dependent on having aws-sam-cli locally. AWS has a doc page an aws-sam-cli with the installation instructions at this url: aws-sam-cli installation

To install in the same environment as the poetry on your system:

pip install poetry
poetry self add poetry-aws-sam

This will add sam as a command in poetry. If you run poetry list, you will see sam as an option amongst the commands. You can then call poetry sam --help for options in running sam in your poetry project.

The plugin replaces the sam build command.

Sam deployment steps

sam build
sam package
sam deploy

Deployment steps using poetry-aws-sam:

poetry sam --without-hashes
sam package
sam deploy

History

The plugin was created while Pinnacle Solutions Group was working with one of its clients on a projects built on aws serverless resources. The client used SAM for its serverless management. The company ended a vendor's contract, which led to the team not having a repository for its package whl's. The team shifted to using poetry and installing its packages from a github url. With the change to poetry, the team needed to update how it interacted with SAM. This plugin was created by Pinnacle Solutions Group to meet that need.

Working on poetry-aws-sam

Common Actions - justfile

There is a justfile for just commands. It is like make with improvements. You can read about installation and usage here: just - github

• show all the commands by running only just: just
• set up git hooks through the pre-commit project: just pre-commit-setup
• run the unit tests: just tests

Install locally to test iteratively

python -m venv .venv
source .venv/bin/activate
pip install -e .
just sam # same as: .venv/bin/poetry sam

This will set up a virtual env and have the sam ready as a plugin

pypi deployment

The master.yml workflow will deploy to pypi. This means that the project version in pyproject.toml will need to be increased when merging into the master branch.

A good dev flow:

• create new branch
• update version number of project
  • can use poetry version minor to increase the minor version
• with code update, create pull request to merge into master
• approve pull request and then merge into master
• the workflow will then deploy to pypi