Smarter Python Configs: an OOP Framework

Managing configurations across multiple environments and services can be a complex task. Developers often face redundant work due to the need for separate config setups for each environment and service. This is where an OOP-based configuration framework in Python comes into play. The framework, built on Pydantic, simplifies and centralizes configuration management.

The configuration structure consists of four layers. The base layer contains shared, global settings like environment identifier, logging level, and GCP credentials. Environment-specific mixins, such as ProdConfigsMixin, override these shared values and add environment-specific configurations. The service layer contains service-specific variables and behavior, while the instance layer includes instance-specific configurations.

By defining configurations as Python classes and using object-oriented programming, the framework encourages composition and reuse. A set of CLI tools, such as ac, allows developers to easily manage and deploy configurations. The ac command can output all resolved configs for the current environment, enabling dynamic generation of environment variables in shell sessions or pipelines. This eliminates the need for handcrafting .env files for CI/CD pipelines, as environment-specific values can be dynamically generated.

The framework's key advantage lies in its ability to encode environment-specific behavior within the config object. CI/CD pipelines can simply run deploy and assert_deploy commands in the appropriate environment. Adding new services or environments requires only the creation of a new config class, rather than a complete overhaul of the pipeline. This results in a more efficient and streamlined process, minimizing redundancy and improving overall delivery speed.