emilosman/monkey

A Go implementation of an API Test CRM service.

monkey

A Go implementation of an API Test CRM service.

Run command

go run monkey.go

Purpose of this exercise

• Learn Go
• Implement an API from scratch
• Get to know the popular gin web framework
• Set up authorization with OAuth 2.0
• Exchange auth token with Google
• Check if the ID token is a valid JWT
• Avoid ORM libraries like GORM
• Write raw SQL for schemas and queries
• Get to know sqlc to compile SQL to type-safe code
• Set up postgres running in podman
• Use golang-migrate to generate migrations

Database setup

1. SQL schema defines the Customer and User models
2. query.sql defines the model relations
3. sqlc.yml config file references the schema and query, along with the database engine and output files that will be generated
4. sqlc generates Go structs that will be used in the Go program to interface with the database. These structs are type safe and protect against SQL injections
5. Various .sh scripts document how to run commands that set up the database
6. Containerized postgres runs on podman machine
7. golang-migrate generates migrations
8. Seed function populates the database initially

References

• sqlc: Getting started with PostgreSQL
• golang-migrate: PostgreSQL tutorial for beginners

REST API

1. gin for handling routes
2. Handlers specify User and Customer routes
3. Model structs are used in route actions, as a form of ORM
4. Queries and auth middleware added to gin context

References

• gin: middleware examples
• gin: Developing a RESTful API with Go and Gin

Authorization

1. Set up project on Google Cloud Console, get client id and secret, set as ENV variables
2. Inject auth middleware into gin context for specific routes
3. Auth route handler handles the callback after authorization, gets id token and sets it in a cookie
4. Auth middleware validates the id token and sets User as admin

References

• OAuth2 for Go
• idtoken for Go
• Access token vs ID token
• JWT debugger
• fiber: OAuth 2.0 Implementation