OSvariables_stories

US IO : Provide environment variables and system-level configurations for portability across various environments.

• US IO : Provide environment variables and system-level configurations for portability across various environments.
  • classes relations
  • User Story: Environment Configuration Management with Singleton and Pydantic
    • 1. User Story: Centralized Environment Configuration
    • 2. User Story: Ensure Consistent MLflow Configuration
  • Common Acceptance Criteria
  • Example Use Cases
  • Definition of Done (DoD):
  • Code location
  • Test location

---

classes relations

classDiagram
    class Singleton {
        - _instances: dict
        + __new__(cls, *args, **kwargs) Singleton
    }

    class BaseSettings {
        <<external>>
    }

    class Env {
        + mlflow_tracking_uri: str = "./mlruns"
        + mlflow_registry_uri: str = "./mlruns"
        + mlflow_experiment_name: str = "regression_model_template"
        + mlflow_registered_model_name: str = "regression_model_template"
    }

    class Config {
        + case_sensitive: bool = False
        + env_file: str = ".env"
        + env_file_encoding: str = "utf-8"
        + extra: str = "ignore"
    }

    Env --|> Singleton : inherits
    Env --|> BaseSettings : inherits
    Env *-- Config : inner class

Loading

User Story: Environment Configuration Management with Singleton and Pydantic

---

1. User Story: Centralized Environment Configuration

Title:
As a developer, I want a centralized and reusable way to manage environment variables and configuration settings so that I can ensure consistent behavior across the application.

Description:
The Env class combines the Singleton pattern and Pydantic's BaseSettings to create a single shared instance for managing environment-specific configurations. It ensures that environment variables are accessible, validated, and loaded from a .env file if available.

Acceptance Criteria:

• The Env class:
  • Inherits from both Singleton and Pydantic's BaseSettings.
  • Ensures only one instance of the class exists throughout the application (Singleton behavior).
  • Provides default values for MLflow-related environment variables.
  • Supports loading configurations from a .env file.
• Environment variables are:
  • Case-insensitive (optional feature).
  • Readable from both the system environment and a .env file.
  • Validated using Pydantic's type system.
• The Config class:
  • Enables .env file reading with UTF-8 encoding.
  • Handles case-insensitive environment variable lookups if enabled.

---

2. User Story: Ensure Consistent MLflow Configuration

Title:
As a machine learning engineer, I want to configure MLflow settings (tracking URI, registry URI, experiment name, etc.) in a single place so that I can easily manage and update them without code duplication.

Description:
The Env class provides default and customizable values for MLflow-related configurations. It ensures that all MLflow components use the same settings consistently, reducing potential misconfigurations.

Acceptance Criteria:

• The following MLflow configuration keys are available:
  • mlflow_tracking_uri: Default is "./mlruns".
  • mlflow_registry_uri: Default is "./mlruns".
  • mlflow_experiment_name: Default is "regression_model_template".
  • mlflow_registered_model_name: Default is "regression_model_template".
• Values can be overridden by environment variables or a .env file.
• The Singleton pattern ensures no accidental re-creation of the Env instance.

---

Common Acceptance Criteria

1. Implementation Requirements:

   • The Singleton pattern ensures a single shared instance of the Env class.
   • Default values are provided for MLflow configuration keys.

2. Environment Variable Management:

   • Values are first read from the environment variables.
   • If not set, values are read from the .env file.
   • Defaults are used if neither source provides a value.

3. Error Handling:

   • The Env class validates all settings using Pydantic's type system.
   • Errors are raised if a setting does not meet its type requirements.

4. Testing:

   • Unit tests verify:
     • Singleton behavior of the Env class.
     • Correct loading of environment variables and .env file values.
     • Validation of MLflow-related settings.
   • Mocking is used to simulate environment variables and .env file contents.

5. Documentation:

   • Class-level docstrings explain the purpose and usage of Env and Singleton.
   • Examples demonstrate:
     • Accessing MLflow settings via the Env class.
     • Overriding values with environment variables or a .env file.

---

Example Use Cases

1. Accessing MLflow Settings:

   env = Env()
   print(env.mlflow_tracking_uri)  # Output: "./mlruns" (default or overridden)

2. Using the Singleton Pattern:

   env1 = Env()
   env2 = Env()
   assert env1 is env2  # True, as both point to the same instance.

3. Overriding via Environment Variables or .env:

   • Set MLFLOW_TRACKING_URI="http://localhost:5000" in .env file or environment.
   • Access with:

     env = Env()
     print(env.mlflow_tracking_uri)  # Output: "http://localhost:5000"

---

Definition of Done (DoD):

• The Env class implements environment configuration management with validation.
• The Singleton behavior is verified, ensuring a single instance is used.
• The .env file and environment variables are correctly loaded and take precedence over defaults.
• All functionality is tested, with error scenarios validated.
• Documentation provides clear usage instructions and examples.

Code location

src/regression_model_template/io/osvariables.py

Test location

tests/io/test_osvariables.py