ymlmaster 0.0.2.1

Powerful and clean YAML-based configuration library for Python 3.11+

ymlmaster

ymlmaster is a configuration loading utility for Python 3.11+. It provides a unified interface for loading deeply structured YAML configuration files (optionally combined with .env overrides) into statically typed Python classes using either dataclass or pydantic.

Features

• Schema generation from YAML into @dataclass or Pydantic models
• Nested structures supported automatically
• Optional .env merging with environment variable fallback
• Profile support (e.g., dev, release)
• CLI generator for schema models

Installation

Installation is done in the project using pip or poetry:

pip install ymlmaster

poetry add ymlmaster

Model Generation CLI

To generate data models from a YAML configuration file:

poetry run generate-schema
  --settings settings.yml
  --output settings_model.py
  --type dataclass # or --type pydantic,  default: dataclass
  --profile dev # or custom name block,  default: dev
  --url-fields postgresql redis # added url services

This will generate Python code like:

@dataclass
class Postgresql:
    host: Optional[str] = None
    user: Optional[str] = None
    password: Optional[str] = None
    port: Optional[str] = None
    db: Optional[str] = None

@dataclass
class Redis:
    host: Optional[str] = None
    port: Optional[str] = None

@dataclass
class Application:
    token: Optional[str] = None
    admin_id: Optional[str] = None

@dataclass
class Settings:
    postgresql: Postgresql = None
    redis: Redis = None
    application: Application = None
    postgresql_url: Optional[str] = None
    redis_url: Optional[str] = None

Using the Loader

You can then load values from settings.yml and .env into your model:

from pathlib import Path
from ymlmaster import SettingsLoader
from <settings-model> import Settings

loader = SettingsLoader(
    settings_path=Path("settings.yml"),
    env_path=Path(".env"),
    model_class=Settings,
    use_release=False, # true - release block, false - dev
    profile=None, # specify the exact loading block
    url_templates={
        "postgresql": "postgresql+asyncpg",
        "redis": "redis",
        "nats": "nats"
    } # url generation instructions   <block name>:<circuit name>
)

config = loader.load()

print(config.redis.host)
print(config.application.admin_id)

Environment Variable Override Behavior

• Values from .env are injected only if the YAML value is null
• Nested overrides use __ as separator:
  • For application.token → APPLICATION__TOKEN
  • For redis.port → REDIS__PORT

If the key ends with port and is an integer, a default IP of 127.0.0.1: is prepended unless already present.

This separation helps to use sensitive data in Dockerfile/DockerCompose and in your project at once Example docker-compose.yml:

services:
  q3s2j0pj0fuj:
    image: "postgres:17"
    container_name: "q3s2j0pj0fuj"
    environment:
      POSTGRES_DB: ${POSTGRESQL__DB}
      POSTGRES_USER: ${POSTGRESQL__USER}
      POSTGRES_PASSWORD: ${POSTGRESQL__PASSWORD}
    ports:
      - "${POSTGRESQL__PORT}:5432"
    logging:
      driver: "json-file"
      options:
        max-size: "10m"
        max-file: "3"
    networks:
      - Network
    restart: always

MIT LICENSE