fragua 0.4.0

Package for creating ETL environments for data analysis

Fragua

Fragua is a lightweight and modular library designed to build ETL/ELT pipelines and data processing workflows in Python. It provides reusable components such as environments, agents, styles, parameters, and storages to orchestrate data extraction, transformation, and loading with traceability and best practices.

---

What is Fragua?

Fragua provides an abstraction layer for data integration tasks based on three main agents:

• Extractor: retrieves data from different sources such as Excel, CSV, or APIs.
• Transformer: transforms or enriches data by applying rules or models.
• Loader: saves or delivers results to final destinations such as files or databases.

In addition to the agents, Fragua introduces the concept of Environments, which act as isolated workspaces that organize and group all components involved in a pipeline. An Environment manages:

• Registered agents and their lifecycle.
• Registries for styles, functions, and params.
• A dedicated Warehouse for storing intermediate or final artifacts.
• Execution settings, logging configuration, and context boundaries.

This design allows you to run multiple pipelines independently, each with its own configuration, state, and stored data.

Fragua also includes a storage system with:

• Warehouse and WarehouseManager to store intermediate artifacts with metadata and traceability.
• A modular architecture where styles, functions, and params can be registered within an Environment.

---

Key Features

• Environment modeling (Environment) to isolate and organize working instances.
• Agents (Extractor, Transformer, Loader) with a common pipeline, undo capability, and operation logging.
• Registries for params, functions, and styles.
• Storage types (Storage, Box, Container) and a centralized Warehouse.
• Built-in utilities for logging, metrics, and execution state summaries.

---

Project Structure

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

---

Installation

Install Fragua in editable mode from the root of the repository:

python -m pip install -e .

Check requirements.txt for additional dependencies.

---

Usage Example

import fragua as fg
from pathlib import Path

BASE_DIR = Path(__file__).parent
INPUT_XLSX = BASE_DIR / "test_files" / "input_files" / "test_data.xlsx"
OUTPUT_XLSX = BASE_DIR / "test_files" / "output_files"

env = fg.create_fragua("fragua_1", "minimal", True)
env.create_extractor("extractor")
env.create_transformer("transformer")
env.create_loader("loader")

extractor = env.get_agent("extractor")
transformer = env.get_agent("transformer")
loader = env.get_agent("loader")

extractor.work(
    "excel",
    save_as="extracted_data",
    path="./test_files/input_files/test_data.xlsx",
    sheet_name=0,
)

transformer.work(
    style="report",
    apply_to="extracted_data",
    save_as="transformed_data",
)

loader.work(
    style="excel",
    apply_to=["extracted_data", "transformed_data"],
    destination="./test_files/output_files",
    file_name="output_file",
)

---

Author

Santiago Lanz
📍 Developer and creator of Fragua
🌐 Portfolio: https://sagodev.github.io/Portfolio-Web-Santiago-Lanz/\ 💼 LinkedIn: https://www.linkedin.com/in/santiagolanz/\ 🐙 GitHub: https://github.com/SagoDev

---

⚖️ License

This project is distributed under the MIT license.
See the LICENSE file for more details.