Infonetica - Configurable Workflow Engine

This project is a take-home exercise to implement a minimal backend service for a configurable workflow engine based on state machines. The API allows clients to define workflows, start instances, and transition them between states by executing actions.

Technical Stack

• Language/Framework: C# / .NET 8
• API Style: ASP.NET Core Minimal APIs
• Persistence: Simple in-memory dictionaries. No database is required.
• Dependencies: The project uses the default ASP.NET Core web SDK and Swashbuckle.AspNetCore for API documentation.

How to Run the Application

Prerequisites

• .NET 8 SDK

Steps

1. Clone the repository:

   git clone <your-repo-url>
   cd Infonetica.WorkflowEngine

2. Ensure Development Environment is Set: This project uses Properties/launchSettings.json to automatically configure the application to run in Development mode. This ensures the Swagger API page is available.

3. Run the application:

   dotnet run

4. Access the API:

   • The application will start and listen on http://localhost:5000.
   • You can explore and interact with the API using the built-in Swagger UI at: http://localhost:5000/swagger

Assumptions and Design Notes

• Persistence: Data is stored in-memory using ConcurrentDictionary. The data will be lost when the application restarts, as allowed by the requirements.
• IDs: WorkflowDefinition and WorkflowInstance IDs are generated as GUIDs for simplicity. State and Action IDs are expected to be meaningful strings provided by the client.
• Validation: Validation is performed within the WorkflowService. It rejects invalid definitions and invalid action executions.
• Known Limitations: The assignment was time-boxed to approximately 2 hours. As such, the code includes // TODO comments for potential improvements that would be addressed with more time.

API Usage Example (via Swagger)

1. Create a Workflow Definition:

   • Execute POST /api/workflows/definitions
   • Provide the following JSON in the request body:

     {
       "id": "leave-request",
       "states": [
         { "id": "draft", "name": "Draft", "isInitial": true },
         { "id": "submitted", "name": "Submitted for Approval" },
         { "id": "approved", "name": "Approved", "isFinal": true },
         { "id": "rejected", "name": "Rejected", "isFinal": true }
       ],
       "actions": [
         {
           "id": "submit",
           "name": "Submit",
           "fromStates": ["draft"],
           "toState": "submitted"
         },
         {
           "id": "approve",
           "name": "Approve",
           "fromStates": ["submitted"],
           "toState": "approved"
         },
         {
           "id": "reject",
           "name": "Reject",
           "fromStates": ["submitted"],
           "toState": "rejected"
         }
       ]
     }

2. Start a Workflow Instance:

   • Execute POST /api/workflows/instances
   • Body: {"definitionId": "leave-request"}
   • This returns a new instance in the draft state. Note the id from the response.

3. Execute an Action:

   • Execute POST /api/workflows/instances/{instanceId}/actions/{actionId}
   • Use the instance id from the previous step for {instanceId} and "submit" for {actionId}.
   • This moves the instance to the submitted state, and its history is updated.