fragua 0.6.0

Package for creating ETL environments for data analysis

Fragua

Fragua is a lightweight, modular Python library designed to model, configure, and orchestrate ETL / ELT workflows through explicit environments and well-defined execution boundaries.

Instead of centering the design around individual agents, Fragua treats the Environment as the primary abstraction: a self-contained execution context that owns configuration, state, components, and data artifacts.

---

What is Fragua?

Fragua provides an abstraction layer for data integration workflows where everything happens inside an Environment.

An Environment represents a fully isolated workspace that defines:

• Which agents exist
• Which components they can use
• Where data is stored
• How operations are logged and traced

Agents such as extractors, transformers, and loaders do not operate independently; they are created, configured, and executed within an Environment, which guarantees consistency, traceability, and predictable behavior.

---

The Environment as the Core Concept

The Environment is the central coordinator of the system.

It is responsible for:

• Managing the lifecycle of agents.
• Hosting registries for:
  • params
  • functions
  • styles
• Providing access to a shared Warehouse.
• Enforcing execution context, configuration rules, and boundaries.
• Exposing structured summaries of the entire runtime state.

This approach allows multiple environments to coexist independently, each one representing a distinct pipeline, experiment, or execution context.

---

Agents as Environment-Orchestrated Components

Within an Environment, Fragua supports three agent roles:

• Extractor — Retrieves data from external sources.
• Transformer — Applies transformations or enrichment logic.
• Loader — Persists or delivers processed data.

Agents are not global actors. They:

• Are instantiated by the Environment
• Resolve styles, params, and functions exclusively via registries
• Interact with data only through the Environment’s Warehouse
• Follow a standardized execution workflow

This ensures that agents remain stateless orchestrators rather than owners of configuration or data.

---

Component Registries and Architecture

Fragua uses a layered component model to ensure consistency and extensibility:

• FraguaComponent — Base abstraction for all registrable elements.
• FraguaSet — Generic container responsible for CRUD operations, validation, and summaries.
• FraguaRegistry — Groups related FraguaSet instances by action and component type.

All components (styles, functions, params, agents) are registered and resolved through these layers, eliminating implicit coupling and duplicated logic.

---

Storage, Warehouse, and Traceability

Each Environment owns a Warehouse, which acts as the single source of truth for pipeline artifacts.

Storage management provides:

• Controlled add, get, delete, rename, and copy operations
• Full movement logging with timestamps
• Undo support for destructive actions
• Metadata-based search and snapshotting

---

Key Features

• Environment-centric workflow orchestration.
• Explicit isolation between pipelines.
• Standardized agent execution model.
• Component-based architecture.
• Centralized storage and traceability.
• Rich summary() methods with improved docstrings.
• Built-in logging and operational metadata.

---

Project Structure

fragua/
├── core/
├── environments/
├── extract/
├── transform/
├── load/
├── utils/
└── __init__.py

---

Installation

python -m pip install -e .

⚠️ Fragua is published on PyPI for learning purposes.
It is not recommended for production use yet.

---

Author

Santiago Lanz
🌐 https://sagodev.github.io/Portfolio-Web-Santiago-Lanz/
💼 https://www.linkedin.com/in/santiagolanz/
🐙 https://github.com/SagoDev

---

License

MIT License