README.md

Functions

This directory contains the functions referenced by the state machine. Each is a self-contained project, which you can open, run, and run tests for with VSCode directly.

Alternatively, you can open workspace.code-workspace and see all four functions at once. With the workspace open, you can run code / run tests on that code for each of the four current functions - the workspace will automatically pick up on the virtual environments referenced in the poetry files.

Dependencies

Each function uses Poetry to manage dependencies, run tests, and package. Unfortunately SAM requires a requirements.txt file. The pyproject.toml files in each function directory is meant to be the source of truth; the requirements.txt files are generated from this.

To create a requirements.txt file, you can do :

poetry export --without-hashes -f requirements.txt -o requirements.txt --with-credentials

You can alternatively run make build.requirements from the top-level directory to create all requirements.txt files at once.

Note: you should not manually alter the requirements.txt file directly. Instead, run poetry add X to add dependency X to your project, then run the above export or make comamnds! This is why requirements.txt has been added to the .gitignore file.

Build

You can build each service by cding into the service directory, then running poetry install, which will create a virtual environment at ./.venv and install all required dependencies. Alternatively do this for all targets at the top level with the makefile target make build.local

Run

You can run each service by cding into the service directory, then running poetry run python app.py, after having run poetry install as above

Test

You can test each service by cding into the service directory, then running poetry run pytest --capture=no --verbose --cov=app --cov-report term-missing. Alternatively do this for all targets at the top level with the makefile target make test.local

Deploy

To deploy infrastructure in this project:

1. From the top level directory, run make deploy.guided the first time, then make deploy subsequent times. Under the hood this will run sam build.requirements to create the requirements.txt files, and sam build to build the local aws project, then deploy it.
2. As part of sam deploy.guided, you will be prompted to add a value for the ENCV_API_KEY parameter - paste your value from encv.org when prompted
3. Important! The project creates a secrets manager storage, but does not populate the GOOGLECREDENTIAL secret itself. You will need to go to SecretsManager and paste the contents of your secrets.json file from the service account associated with your GCP project into the secret with a name matching the one you choose to supply as GOOGLE_APPLICATION_CREDENTIALS_SECRET.