Premade packages to run on aws lambda layers
The docker-compose includes a service that simply runs the CLI. The requirements are installed by pip, so whatever pip recognizes will be accepted by the script. To create a layer with numpy+pandas you would simply run
docker-compose run make numpy pandas --name NumPandas --desc "numpy and pandas"
To get help,
docker-compose run make --help
Which will print (might change if I forget to update readme)
Usage: make_layer.py [OPTIONS] [REQUIREMENTS]...
Options:
--name TEXT Name of the resulting layer. [required]
--desc TEXT Description for lambda layer [required]
--help Show this message and exit.
If no requirements are specified, the script will attempt
to pip install from the ./requirements.txt
file which
is mounted to /code/requirements.txt
which in the default
docker compose is mounted from the current directory.
So if nothing is changed, this will install the
requirements.txt from the current directory.