A DSL to define and enforce governance policies

This repository contains the definition of a domain-specific language (DSL) for governance policies (the engine for the enforcement of the policies is available here). We provide details below for the contents of this repository. The DSL provides a form-based editor, available here.

Contents

The repository structure is divided in five folders:

• governancedsl/grammar/: This folder contains the definition of the grammar, or concrete syntax, of our DSL. We used ANTLR to define our concrete syntax, see govdsl.g4 for the implementation.
• governancedsl/metamodel/: This folder contains the metamodel of the abstract syntax of our DSL:
  • metamodel.py: The implementation of the abstract syntax metamodel as Python classes.
  • ocl_constraints.ocl: Set of OCL constraints applied to the metamodel (OCL constraints are also included in the BESSER diagrams).
  • Since the metamodel has been created using the BESSER Web Modeling Editor we provide several files under metamodel/besser_models/ folder:
    • .json files: The JSON representation that can be used to open the metamodel in the editor.
    • .svg and .png files: The SVG and PNG representation for visual purposes.
    • .drawio.png files: We also provide the metamodel embedded in a PNG file that can be opened in the draw.io editor.
• governancedsl/utils/: This folder contains additional support for the definitions of exceptions, extensions of our metamodel (e.g., code-hosting platform extension), and data structure support for the parser.
• UI/: This folder contains a Gradio-based form editor for creating governance policies interactively:
  • See UI/README.md for detailed information.
  • The form-based editor is available here.
• tests/: This folder contains the tests. There are three subfolders inside the test_cases/ for the examples:
  • invalid_examples/: Here we define with our DSL different invalid policies (e.g., the required number of votes is negative).
  • NL_examples: Here we define the examples in natural language, which can come from existing repositories (NL_examples/real-world/ folder) or created from us (NL_examples/artifical/).
  • valid_examples: Here we define with our DSL different valid policies, which can also come from NL_examples/.

Tests

When contributing to incorporate new tests in natural language (NL_examples/) we recommend using the following template:

Original repo: URL (or we can also report "owner/name" if repository is from GitHub)
Access date: Date you accessed the repository
Reporter: -
Policy: If multiple, enter a new line for each.

Prerequisite

• Python 3.11
• Recommended: Create a virtual environment (e.g. venv, conda)
• Clone this repository
• Install the dependencies by referencing to the requirements files:

pip install -r requirements.txt

Run Tests

From the project root directory, run all tests with:

python -m unittest tests.test_policy_creation -v

Or run a specific test:

python -m unittest tests.test_policy_creation.testPolicyCreation.test_majority_policy_creation -v

Note: Make sure you're in the project root directory when running the tests to ensure proper module imports.