Serverless Python Requirements

A Serverless v1.0 plugin to automatically bundle dependencies from requirements.txt and make them available in your PYTHONPATH.

Install

npm install --save serverless-python-requirements

Add the plugin to your serverless.yml:

plugins:
  - serverless-python-requirements

Adding the dependencies to sys.path

Automatic

The default behavior of this plugin is to link libraries into the working tree during deployment so that they are in your handler's PYTHONPATH when running on lambda.

Manual

This method is required when using ZipImport support and can be enabled manually by adding the following option to your config:

custom:
  pythonRequirements:
    link: false

serverless-python-requirements adds a module called requirements to your puck. To easily make the bundled dependencies available, simply import it. Eg. add this to the top of any file using dependencies specified in your requirements.txt:

import requirements
# Now you can use deps you specified in requirements.txt!
import requests

Cross compiling!

Compiling non-pure-Python modules is supported on MacOS via the use of Docker and the docker-lambda image. To enable docker usage, add the following to your serverless.yml:

custom:
  pythonRequirements:
    dockerizePip: true

ZipImport!

To help deal with potentially large dependencies (for example: numpy, scipy and scikit-learn) there is support for having python import using zipimport. To enable this add the following to your serverless.yml:

custom:
  pythonRequirements:
    zipImport: true

Limitations

• if using the package directive in serverless.yml ensure that requirements.py is are included as well as .requirements or .requirements.zip if using ZipImport.

Manual invocations

The .requirements and requirements.py files are left behind to simplify development. To clean them up, run sls requirements clean. You can also install them manually for local development with sls requirements install.

Credit

This plugin is influenced by serverless-wsgi from @logandk. I however wanted a simpler pip install process. It now also supports bundling packages without the wsgi handler.