BDDTestSuite is a comprehensive end-to-end testing framework for an e-commerce web application, built on .NET 8 with Selenium WebDriver, Reqnroll (SpecFlow) and xUnit. Following a Behavior-Driven Development (BDD) workflow, it uses Gherkin feature files to describe user journeys and step definitions to drive the automation. Integrates Google’s Gemini API to deliver AI-driven analysis, root-cause detection and natural-language summaries of your test results.
- Project Structure
- Configuration Management
- Logging
- Utility Classes
- AI-Powered Analysis, Root Cause Detection & Summaries
- Reporting
- Scenarios & Features
BDDTestSuite/
├── Features/ # Gherkin feature files describing scenarios
├── Models/ # Data models (e.g., Product)
├── PageObjects/ # Page Object Model classes for UI abstraction
├── StepDefinitions/ # Step definitions mapping Gherkin steps to code
├── Utils/ # Utility/helper classes (browser, waits, etc.)
├── config.json # Test configuration (browser, credentials, etc.)
├── Hooks.cs # Test lifecycle hooks (setup/teardown)
Configuration is managed via config.json which contains:
- The base URL of the application
- Browser selection (
chrome,firefox,edge) - Test user credentials (standard, locked-out, etc.)
Example:
{
"url": "<https://www.saucedemo.com/>",
"browser": "firefox",
"username": {
"standard-user": "standard_user",
"locked-out-user": "locked_out_user"
},
"password": {
"valid": "secret_sauce",
"invalid": "invalid_pass"
}
}
Configuration is loaded at test startup in Hooks using Microsoft.Extensions.Configuration.
IConfigurationRoot configuration = new ConfigurationBuilder()
.AddJsonFile("config.json")
.AddEnvironmentVariables()
.Build();
string browser = configuration["browser"];
Logging is implemented using Serilog. Logs are written both to the console and to a file (log.txt). Each scenario's log entries are tagged with the scenario name for traceability.
- Log output includes timestamps, log level, scenario name, and message.
- Logging is initialized in
Hooks. - Each step and page object method logs key actions and errors.
Example log entry:
[18:27:10 INF] [Scenario: Sort products in ascending order by price] Navigating to the login page: https://www.saucedemo.com/
Utility classes in Utils/ provide reusable helpers:
BrowserUtil: Browser initialization and navigation.WaitUtils: Explicit waits for elements (visible, clickable, URL changes, etc.).ElementUtils: JavaScript-based element interactions (e.g., JS click).ReportUtils: Step node creation and failure logging.ScreenshotUtils: A utility class that captures full‑page browser screenshots and returns them as Base64‑encoded strings.
These utilities help keep step definitions and page objects clean and robust.
-
Automated Test Result Summaries
- After each test run, the suite calls Gemini to generate a concise, natural-language overview of failed scenarios.
-
Insight Extraction
- Leveraging feature names, scenario steps, logs, and screenshots, Gemini extracts root causes, identifies common failure patterns, and pinpoints potential sources of test flakiness.
-
ExtentReports Integration
- The generated summary is seamlessly appended to the scenario's failed step in the report, embedding AI-driven insights directly alongside your test results.
- A snapshot of the report highlighting a failed inventory test, where the AI correctly pinpoints an incorrect element locator as the root cause of the failure.
- A snapshot of the report showing a failed checkout form test, where the AI pinpoints that the test bypassed the checkout form page and proceeded to the subsequent page.
This automation test suite integrates Extent Reports and Serilog to give you rich, step‑by‑step insights into every test run:
-
Extent Reports
- Generates an HTML report for each test run.
- Automatically attaches a screenshot whenever a test fails, making it easy to see exactly what went wrong.
-
Serilog Sink for In‑Test Logging
- A custom Serilog sink captures all log events emitted during a test.
- Log events are queued per test execution and then dequeued when attaching to Extent Report nodes.
- Each BDD step node in the report includes its own detailed log output, so you can trace the exact sequence of actions and messages.
-
AI Scenario Summary Integration
- The generated summary is seamlessly appended to the scenario's failed step in the report, embedding AI-driven insights directly alongside your test results.
Scenarios are defined in Gherkin feature files under Features/:
- Login (Login.feature): Valid/invalid login, locked-out user, empty credentials.
- Inventory (Inventory.feature): Viewing product details, sorting products by name/price.
- Cart (Cart.feature): Adding/removing products, verifying cart contents.
- Checkout (Checkout.feature): Form validation, overview page, completing purchases.
Each scenario is mapped to step definitions in StepDefinitions/, which use page objects and utilities to interact with the application.