Self-Heal Framework (SeleniumWire + Python)

Overview

This repository contains a custom Selenium automation framework with a basic self‑healing mechanism, built using Python + Pytest.

The main goal of this project is to explore how modern automation tools handle locator failures and to implement a lightweight, transparent self‑healing approach without relying on commercial tools.

This is not a wrapper around Selenium and not a toy project. It focuses on:

• Practical locator recovery
• Maintainable structure
• Clear separation of responsibilities

Why Self‑Healing?

In real projects, UI tests fail frequently due to:

• Minor DOM changes
• Updated attributes (id, name, xpath)
• UI refactoring without functional changes

Instead of immediately failing the test, this framework:

1. Tries the primary locator
2. Falls back to alternative locators
3. Updates the locator source automatically (if enabled)

This reduces flaky failures and improves test stability.

Key Features

• JSON‑based locator management (primary + fallback)
• Centralized self‑healing engine
• Automatic fallback resolution
• Custom logging (no print statements)
• Pytest‑based execution
• Page‑Object‑Model friendly structure

Project Structure

Self_Heal_Framework/
│
├── core/
│   ├── self_healing_engine.py   # Core healing logic
│   └── smart_locator.py         # Locator resolution & fallback handling
│
├── network_utils/
│   └──network_utility.py        # Capture and format the APi during the execution
│
├── utils/
│   ├── json_operations.py       # JSON read/write utilities
│   └── logger.py                # Custom logging configuration
│
├── locators/
│   └── login_page.json          # Primary + fallback locators
│
├── tests/
│   └── test_login.py            # Sample test cases
│
├── conftest.py                  # Pytest setup
├── requirements.txt
└── README.md

Locator Strategy (JSON Driven)

Each element is defined with:

• One primary locator
• One or more fallback locators

Example:

{
  "username": {
    "primary": {"type": "id", "value": "user-name"},
    "fallback": [
      {"type": "xpath", "value": "//input[@name='user-name']"},
      {"type": "css", "value": "input[data-test='username']"}
    ]
  }
}

The framework always prefers the primary locator and switches to fallback only on failure.

How Self‑Healing Works (High Level)

1. Test requests an element by logical name (not Selenium locator)

2. Smart Locator fetches primary + fallback locators from JSON

3. Self‑Healing Engine tries each locator in order

4. On success:

   • Element is returned
   • (Optional) locator source is updated

5. On failure:

   • Clear error with logging context

This keeps test scripts clean and readable.

How to Run

Prerequisites

• Python 3.9+
• Chrome browser
• ChromeDriver (compatible version)

Setup

pip install -r requirements.txt

Execute Tests

pytest -v

Logging

• Uses Python logging module
• Logs locator attempts and healing decisions
• Designed to integrate with reporting tools (Allure planned)

Network Utilities

Captures and filters backend API calls during UI execution.
Used by self-heal logic to distinguish UI vs backend failures.

Current Status

• Core self‑healing logic implemented
• JSON‑based locator strategy complete
• Sample test cases added
• Reporting integration planned

Disclaimer

This project is built as a learning‑focused engineering exercise, inspired by tools like Testim, Mabl, and Healenium.

The intention is to understand the mechanics behind self‑healing, not to replace enterprise‑grade tools.

Author

Prasad Helaskar

Automation Engineer | Python | Selenium | Pytest

Future Enhancements

• Locator reliability scoring
• Healing confidence threshold
• Allure reporting integration
• Parallel execution support

Note

This framework is actively evolving. Test coverage and locator strategies are being expanded incrementally to reflect real-world automation practices.