Skip to content

Configuration Components¤

Purpose¤

The components module provides specialized functional units that handle specific aspects of configuration processing. Each component focuses on a single responsibility, creating a modular and testable architecture.

Design Rationale¤

Components implement the Single Responsibility Principle and are designed for: - Testability: Each component can be tested in isolation - Reusability: Components can be used independently or composed - Maintainability: Clear boundaries between different configuration operations - Extensibility: New components can be added without affecting existing ones

Component Overview¤

graph LR A[UnifiedYAMLLoader] --> B[Configuration Dict] B --> C[ExperimentBuilder] C --> D[Built Configuration] D --> E[UnifiedValidator] E --> F[Validation Result] F --> G[UnifiedMerger] G --> H[Final Configuration]

Key Components¤

Validation Framework (validators.py)¤

  • UnifiedValidator: Orchestrates multi-stage validation
  • BusinessRulesValidator: Domain-specific validation rules
  • CompatibilityValidator: Plugin and version compatibility checks

Configuration Construction (builders.py)¤

  • ExperimentBuilder: Constructs experiment configurations with auto-fix
  • ServiceBuilder: Handles service configuration and port management
  • GlobalConfigBuilder: Processes global settings with environment resolution

Configuration Loading (loaders.py)¤

  • UnifiedYAMLLoader: Advanced YAML parsing with interpolation
  • UnifiedVersionLoader: Version-aware configuration loading
  • PluginConfigLoader: Dynamic plugin configuration discovery

Configuration Merging (merger.py)¤

  • UnifiedMerger: Sophisticated configuration merging
  • Multiple strategies: Replace, append, deep merge with conflict resolution

Entry Points¤

Direct Component Usage¤

from panther.config.core.components.validators import UnifiedValidator
from panther.config.core.components.builders import ExperimentBuilder
from panther.config.core.components.loaders import UnifiedYAMLLoader

# Load configuration
loader = UnifiedYAMLLoader()
config_dict = loader.load_from_file("experiment.yaml")

# Build experiment configuration
builder = ExperimentBuilder()
config = builder.build_experiment_config(config_dict)

# Validate configuration
validator = UnifiedValidator()
result = validator.validate_experiment_config(config)

from panther.config.core.manager import ConfigurationManager

# Components are used internally by the manager
config_manager = ConfigurationManager()
config = config_manager.load_and_validate_config("experiment.yaml")

Component Interactions¤

Validation Pipeline¤

  1. Schema Validation: Pydantic type checking and required fields
  2. Business Rules: Domain-specific constraints and relationships
  3. Compatibility: Plugin version and feature compatibility
  4. Auto-Fix: Intelligent correction of common issues

Building Process¤

  1. Global Config: Process system-wide settings
  2. Test Config: Build individual test configurations
  3. Service Config: Handle service definitions and port assignments
  4. Validation: Ensure configuration consistency

Loading Chain¤

  1. File Reading: YAML/JSON parsing with error handling
  2. Environment Resolution: Variable interpolation and substitution
  3. Plugin Discovery: Dynamic schema contribution
  4. Caching: Performance optimization for repeated loads