arta 0.7.0b2

An Open Source Rules Engine - Make rule handling simple

Make rule handling simple

---

Documentation: https://maif.github.io/arta/home/

Repository: https://github.com/MAIF/arta

---

Overview

Arta is a simple python rules engine designed for python developers.

Goal

There is one main reason for using Arta and it was the main goal of its development at MAIF: increase business rules maintainability.

In other words, facilitate rules handling in our python apps.

Origins

The need of a python rules engine emerged when we were working on a new major release of our internal use of Melusine (i.e., email qualification pipeline with ML capabilities).

We were looking for a python library to centralize, manage and standardize all the implemented business rules we had but didn't find the perfect fit.

Therefore, we decided to create this package and by extension of the MAIF's values, we planned to share it to the community.

Features

• Standardize the definition of a rule. All rules are defined the same way in a unique place.
• Rules are released from the code base, which is less error prone and increases clearness.
• Use Arta whatever your field is.
• Great combination with Machine Learning: groups all the deterministic rules of your ML projects.

A Simple Example

Create the three following files and run the main.py script (python main.py or python3 main.py).

rules.yaml :

---
rules:
  default_rule_set:
    admission:
      ADMITTED:
        simple_condition: input.power=="strength" or input.power=="fly"
        action: set_admission
        action_parameters:
          value: true  
      NOT_ADMITTED:
        simple_condition: null
        action: set_admission
        action_parameters:
          value: false

actions_source_modules:
  - actions

actions.py :

from typing import Any


def set_admission(value: bool, **kwargs: Any) -> dict[str, bool]:
    """Return a dictionary containing the admission result."""
    return {"is_admitted": value}

main.py :

from arta import RulesEngine

eng = RulesEngine(config_path=".")

data = {
        "id": 1,
        "name": "Superman",
        "civilian_name": "Clark Kent",
        "age": None,
        "city": "Metropolis",
        "language": "english",
        "power": "fly",
        "favorite_meal": "Spinach",
        "secret_weakness": "Kryptonite",
        }

result = eng.apply_rules(input_data=data)

print(result)

You should get: {"admission": {"is_admitted": True}}

Check the A Simple Example section for more details.

Installation

Install using pip install -U arta. See the Install section in the documentation for more details.

What's New

Want to see last updates, check the Release Notes or the Changelog.

Community

You can discuss and ask Arta related questions:

• Issue tracker:
• Pull request:

Contributing

Contributions are very welcome!

If you see an issue that you'd like to see fixed, the best way to make it happen is to help out by submitting a pull request implementing it.

Refer to the CONTRIBUTING.md file for more details about the workflow, and general hints on how to prepare your pull request. You can also ask for clarifications or guidance in GitHub issues directly.

License

This project is Open Source and available under the Apache 2 License.