implement user favorite management features

[antonioualex]

GWI Favorites Asset Service - Implementation

Overview

This PR implements a user favorites management system for the GWI platform, allowing users to manage personalized lists of favorite assets (Charts, Insights, and Audiences) through a RESTful API.

🚀 Features Implemented

Core API Endpoints

• GET /favorite/ - Retrieve all user favorites
• POST /favorite/{aid} - Add asset to favorites
• PATCH /favorite/{id} - Update favorite description
• DELETE /favorite/{id} - Remove from favorites

Database & Persistence

• PostgreSQL with Docker containerization
• Migration System for schema management and data seeding
• Repository Pattern for data access abstraction

🔧 Technical Implementation

Technology Stack

• Go 1.24+ with modern best practices
• PostgreSQL for data persistence
• Docker & Docker Compose for containerization
• Testcontainers for integration testing
• Mockgen for test mocking
• Bun ORM for PostgreSQL

Testing Strategy

• Unit Tests: Individual component testing with mocks
• Integration Tests: Database integration with Testcontainers
• Test Coverage: Comprehensive coverage across all layers
• CI/CD: GitHub Actions for automated testing

📁 Project Structure

├── app/                    # Application layer (use cases, business logic)
├── cmd/                    # Application entry points
├── domain/                 # Core domain entities and interfaces
├── infra/                  # Infrastructure setup (Docker, migrations)
├── internal/               # Internal packages (config, server setup)
├── persistence/            # Data access layer (repositories)
└── presentation/           # HTTP handlers and routing

[alesr]

Hi @antonioualex, thanks so much for submitting your solution for the assignment! I’ve left a couple of questions for you to take a look at when you have a moment. No rush, take your time. 😊

[alesr]

Could you please elaborate on what is happening here?

The server returns an error, logs the error and then?

[antonioualex]

The server returns an error, logs the error and then calls cleanExit(1). Inside cleanExit(), os.Exit(1) is scheduled as a deferred function. When cleanExit() finishes, the deferred os.Exit(1) executes immediately. The cleanExit function doesn't actually accomplish what its comment and name suggests creating confusion about when exit will occur. Besides that, there's nothing left in that scope to run. I would remove cleanExit() entirely and use os.Exit(1) directly in main() so it can be clearer and doesn't create false expectations.

[alesr]

Why do we need to pass nil to the WithTLSConfig?

Is TLS enabled by default? What does it mean to add this flag and pass a nil configuration?

[antonioualex]

It means no custom TLS configuration is provided which overrides the default behavior. When I pass nil to WithTLSConfig(), the connector will use disabled ssl mode which is needed for local development. The correct approach would be to use, pgdriver.WithInsecure(cfg bool) and pass the value through configuration.

[alesr]

What would be the benefit of not initializing the database here, but pass it as an argument to the run function?

[antonioualex]

🤔 It would give better testability by giving the opportunity to inject a mock database.

[alesr]

What happens to the insert if the user closes the connection with the server?

[antonioualex]

A context cancellation will occur and if it is before or during the database operation execution, then it will not be performed and an error context.Canceled will be returned. In this case it will also log user info and will return an error InternalServerError in the application layer, which is not correct, should return (and handle) a different type of error. If context cancellation occurs after the database write completes, then the data will be stored in the database and the connection will close before the success response reaches the client.

[antonioualex]

Hi @alesr, thank you very much for your review, appreciate it!! 😊 😄 😄