My RESTful API Boilerplate

Image Copyright © SAWARATSUKI. All rights reserved.
Image used under permission from the copyright holder.

This project provides a boilerplate for building RESTful APIs using Go. It is designed to a quick start with best practices, easy configuration, and a clear project structure.

Warning

This boilerplate is specifically tailored for extensive codebases and scalable RESTful API applications, along with frontend websites that may span multiple sites. Its design ensures scalability and straightforward maintenance, making it highly suitable for complex projects. However, it may not be the ideal choice for smaller-scale RESTful API applications or single-website frontends where such robust architecture is not required.

Features

• Use Fiber Framework
• Pre-configured MySQL and Redis integration
• Middleware examples for logging and authentication
• Scalable project structure
• Dual-licensed under BSD 3-Clause and MIT (specific files)
• High Performance (Thanks to the Sonic JSON serializing & deserializing library)
• Cryptography Techniques
• High Quality Go Codes
• Rich Presence TUIs using charm.sh

Warning

Some features might not work as expected when running this repo on certain cloud providers. For example, the Rich Presence TUIs feature requires a tty. The reason why some features might not work as expected after implementation is because this repo is designed to be top-level and follow the Unix philosophy. Plus, most of the Go code in this repo follows many best practices and idioms from Effective Go.

TODO Features

Move here

Note

Additional features will be added later as my primary focus is on using this with the Fiber framework and Cryptography Techniques.

Supported Additional Features

Note

The following list includes additional features that I've been using before for extremely scalable and high-performance applications:

• Protocol Buffers
• Templ

Motivation

The motivation for sharing this RESTful API Boilerplate on GitHub is to streamline my development process, ensuring I don't have to start from scratch each time. It's a robust, secure foundation that adheres to best practices, designed for quick and efficient project initiation.

Philosophy

This boilerplate is grounded in the Unix philosophy, emphasizing simplicity, modularity, and reusability. Each component is designed to do one thing and do it well, with the ability to combine components seamlessly to perform complex tasks. This approach ensures that the boilerplate remains lightweight, efficient, and easy to maintain.

Resource Memory Usage

• MySQL without Redis + 100K POST Request (Outdated - go1.23.2)

Note

The Resource Memory Usage section demonstrates how Go has stable and low memory overhead compared to other languages, especially Java (See this article for more information on Java memory management Lmao.) hahaha.

• Idle (Outdated - go1.23.2)

Note

The Idle section demonstrates the memory usage when there is no request. The maximum average memory usage is around 21.5MB. Go routines (100 goroutines) along with a semaphore are used to automatically handle high traffic situations (e.g., lots of requests).

Also Note that even with high incoming traffic (e.g., 1 million requests), the maximum average memory usage is still relatively low, around 100MB (e.g., 50MB), due to the use of the semaphore.

Latest

• Idle (Latest Go Version - go1.23.3)

sample#memory_total=8.01MB
sample#memory_rss=7.92MB
sample#memory_cache=0.09MB
sample#memory_swap=0.00MB
sample#memory_pgpgin=0pages
sample#memory_pgpgout=0pages
sample#memory_quota=1024.00MB

Note

Go Memory Usage Improvements:

• The Idle memory usage in the latest version of Go has been optimized, reducing it to approximately 7.92MB compared to the previous version.

Average Memory Usage Breakdown:

• The average memory usage of 11.1MB includes:
  • Go routine scheduler for database operations (separate from the 100 goroutines used for handling high incoming traffic automatically with a semaphore)
  • Load for the front-end website
  • Load for the REST APIs (Full Stack)
• The application remains stable without any performance issues, such as bottlenecks or memory leaks, and has zero vulnerabilities.

Go Performance Comparison:

• Compared to other languages, particularly for full-stack development, Go demonstrates superior performance.

Architecture

Below is the architecture of this boilerplate and how it looks. I created this for REST APIs about volcano 🌋 monitoring used by the government (has been done before), so it can easily monitor volcanoes in the real world.

• Backend (Pure Go)

backend/
|-- cmd/
|   `-- server/ (main application entry point)
|       `-- main.go
|-- pkg/
|   |-- restapis/ (API route handlers)
|   |-- any/ (any related code)
|-- internal/ (private application and library code)
|   |-- any/ (any related code)
|-- .env (optional environment variables since it can placed anywhere)
`-- go.mod (dependencies)

• Frontend (Any Framework TS e.g, React,NextJS or pure JS for HTMX)

Example:

frontend/
|-- pages/
|   |-- api/ (optional, for Next.js API routes if needed)
|   |-- _app.js (global page layouts and state)
|   |-- index.js (home page)
|   `-- [...other pages]
|-- public/ (static files like images, fonts)
|-- src/
|   |-- components/ (shared React components)
|   |   `-- [various components]
|   |-- styles/ (global styles, theme)
|   |-- hooks/ (custom React hooks)
|   |-- utils/ (utility functions)
|   |-- lib/ (libraries and configurations)
|   `-- context/ (React context files for state management)
|-- .env.local (environment variables)
|-- next.config.js (Next.js configuration)
`-- package.json (dependencies and scripts)

License

This project is dual-licensed under the BSD 3-Clause License and the MIT License - see the LICENSE file for details.