This simple Docker container is used to build AWS Lambda deployment ZIP files for Python functions on a Windows machine. This is due to the OS differences when installing some packages (such as pandas).
To run this container, you need the following:
- Docker for Windows running with Linux containers
- An input folder that contains your Lambda function and a
requirements.txtfile - An output folder to export the results
To build a Lambda function, you would run this container with the following command:
docker run -v <input-path>:/build/in:ro -v <output-path>:/build/out surfkansas/lambda-build
The <input-path> value should be the full path to your input folder.
The <output-path> value should be the full path to your input folder.
All files in the <input-path> folder will be recursively added to the zip file. Also, a pip restore will be run on the requirements.txt, with the output folder being the root of the zip.
The resulting zip file will be called lambda-build.zip
For this example, we will assume that out input folder is c:\lambdabuild\in and out output folder is c:\lambdabuild\out.
docker run -v c:/lambdabuild/in:/build/in:ro -v c:/lambdabuild/out:/build/out surfkansas/lambda-build
This will produce a deployable Lambda zip file at c:\lambdabuild\out\lambda-build.zip.