Skip to content

Latest commit

 

History

History
346 lines (235 loc) · 11.8 KB

README.md

File metadata and controls

346 lines (235 loc) · 11.8 KB
Raw

Welcome to Fireplace API

Fireplace API is a fully-featured ASP.NET Core Web API sample, and also a Reddit API clone that has communities, posts, and nested comments.

The API supports many features such as DDD structure, Swagger UI, integration testing, various authentication methods, error handling, logging system, caching system, security features, CICD, and Docker deployment.

Stack: REST API, GraphQL, ASP.NET Core (7.0), PostgreSQL, Redis, Swagger, Nginx, CICD, Docker, GitHub Actions, Google OAuth 2.0, Gmail API, NLog, xUnit





- Following Sections:



The DDD Architecture

This project has been divided into multiple subprojects to implement a domain-driven design structure:

  • Application
  • Domain
  • Infrastructure
  • Integration Tests

Layers:

According to Eric Evans's book Domain Driven Design

The DDD patterns help you understand the complexity of the domain.

  • The Application Layer

    • Defines the jobs the software is supposed to do
    • It interacts with the application layers of other systems
    • It does not contain business rules or knowledge
    • Must not directly depend on any infrastructure framework.

  • The Domain Layer

    • The heart of business software
    • Responsible for representing concepts of the business, information about the business situation, and business rules.
    • Must completely ignore data persistence details.

  • The Infrastructure Layer

    • Defines how the data is persisted in databases or other persistent storage
    • Responsible for connecting to external systems and services
    • It does not contain business rules or knowledge



Summary

1. DDD Architecture

  • Layers as subprojects: Application, Domain, Infrastructure
  • API: Handles the client requests
  • Domain: Handles business rules & logic
  • Infrastructure: Handles data persistence & third-party systems

2. Swagger

  • Check The Swagger UI
  • Has the list of API actions to view or run
  • Has examples of possible responses for each action
  • Supports Bearer Authentication
  • Has a custom Google Sign-In button
  • Automatically save cookies after login
  • Automatically use the CSRF token
  • Has customizations with custom-swagger-ui.css & custom-swagger-ui.js

3. GraphQL

  • Check The GraphQL Playground, which is an interactive and graphical IDE for GraphQL
  • With GraphQL you can prevent over-fetching and under-fetching
  • With GraphQL you can call multiple API actions with just a request
  • With GraphQL playground you can check all the schema you can call
  • Supports nested data such as community posts and post comments
  • Supports self-data such as created posts, created comments, and joined communities

4. Caching System with Redis

  • Getting recent request traces from the database took more than 50ms. So I add Redis to cache them. Now API is much faster.
  • Each model has its own cache service class. Check RequestTraceCacheService
  • Check RedisGateway
  • The project caching system is in its starting phase. There are a lot more items to be cached.

5. CICD with GitHub Action

  • Check The CICD.yml
  • Has three steps: Test, Build, and Deploy
  • At the test stage, integration tests will be run.
  • At the build stage, a docker image will be created and pushed to Docker Hub
  • At the deploy stage, the image will be run in a VPS with docker-compose

6. Docker & Docker Compose

7. Integration Testing

  • With xUnit
  • Each test directory is dedicated to several test classes related to an item
  • Each test directory uses a collection name
  • Each test collection has its own database clone
  • Parallelism is used for each test collection
  • Each test function is isolated

8. Database with PostgreSQL

  • Has a repository for each entity
  • Use Fluent API
  • Use ulong Id
  • Use case-insensitive collation
  • Each entity inherits from a base class that has Id, CreationDate, and ModifiedDate

9. Pagination vs Query Result

  • After implementing both stateful & stateless paginations, I decided to remove them!
    • Stateful paginations need a considerable amount of server resources to be run!
    • Stateless paginations can produce duplicate data and even miss some data of different sorts!
  • Use Query Result instead of Pagination
    • The idea is to send a list of the remaining ids for each query
    • Query result is a collection of items and a list of more item ids
    • Also nested comments have child comments and a list of more child comment ids

10. Supports nested comments

  • Each comment has a field for its replied comments
  • Can get post comments along with their replies for a specific sort with the structure

11. Prevents CSRF attacks

  • The token will be refreshed for every unsafe HTTP method call
  • The token should be equal in both cookies and headers

12. Error Handling

  • The error json has a code, a type, a field, and a message
  • Error code is unique to a pair of error type and error field
  • Error types are ALREADY_EXISTS, INVALID_FORMAT, AUTHENTICATION_FAILED, and etc
  • Error field refers to the item that caused the error
  • Error client messages are stored in the database
  • Error codes can be changed freely in the database without changing the production codes
  • Exception middleware catches ApiExceptions, and the other Exceptions occurred in the code
  • Server Message is for the logging, and Client Message is for the users

13. Has logging system

14. Sign up and Log in methods

  • Use Bearer Scheme
  • Sign Up: Google Sign-in, Email
  • Log In: Google Sign-in, Email, Username
  • Can specify the access token in both cookies and headers
  • Use Gmail API for sending emails

15. Request Tracing & API Request Rate Limit

  • Request details will be stored in the database
    • Method, Action, URL, IP, User Agent, User ID, Duration, Status Code, Error Name, Request Time, …
  • Currently, only allow a few API calls per one hour

16. Supports user sessions

  • Can get the list of sessions and revoke them
  • Currently, sessions only have IP, but I should also add the user-agent and the country of the IP

17. Id Generation and Encoding



The Swagger

With the swagger UI, you can interact with the API and discover it. It shows all routes, inputs, outputs, models, and errors. It also provides a swagger.json, which describes the schema of the API that can be imported to your project.

Check the Swagger UI website










How to run a clone

1. Create an empty PostgreSQL database & get its connection string

The connection string should be something like this:

Host=<server-address>;Port=1234;Username=<username>;Password=<password>;Database=<database-name>;Include Error Detail=true;Log Parameters=true;

2. Clone the project

> git clone https://github.com/soroushkavousi/fireplace-api.git
> cd fireplace-api

3. Set the environment variables


Environment Key Value Default Required
FIREPLACE_API_CONNECTION_STRING <connection-string> - ✔
FIREPLACE_API_LOG_DIRECTORY path/to/logs Project Root ✖
FIREPLACE_API_ASPNETCORE_ENVIRONMENT 'Development' or 'Production' 'Development' ✖

4. Apply the database migrations

> dotnet ef database update --startup-project Application --project Infrastructure

5. Run the projects

> dotnet run --project Application

Now you can check the swagger: http://localhost:5000/swagger

Note: At this stage, you may have noticed that some errors say there are no configs or errors in the database! So we are going to fix them.


6. Import the initial data (configs & errors) into the database

The file Guides/db-initial-data.txt has multiple insert queries to feed the initial data. You have two ways to inject the data into the database:

Option 1: Using psql command

> psql -h localhost -p <port> -U <user> -d <database-name> -f "Guides/db-initial-data.txt"

Option 2: Running queries directly

You can easily run queries directly by copying the file content.


7. Test the project

> dotnet test --logger "console;verbosity=detailed"

Congratulations, you have almost run a clone.


8. Customize the configs data

Global variables and sensitive data are stored in a table in the database. You fed sample config data at step 6, but you need to alter it with your project specifications. For instance, generate and set your own Google client-id and client-secret to use Google OAuth 2.0.