Neurostep / todo

Simple Todo list application

Simple Todo application

This repository contains a simple web application, representing as an HTTP API service for a Simple TODO application.

Features:

• CRUD for TODO item
• Add a comment for the TODO
• Add a label for TODO

You can find an API spec in the api/swagger.yaml

Run the application locally

To simply run this application, all you have to have installed is Docker engine with docker-compose toolset. Once you have it installed, please follow steps:

1. Download this repo
2. run docker-compose build
3. run docker-compose up -d

The application will be exposed on port 19000. So, go to the http://localhost:19000/api/v1/todos to check If the application run correctly, you can start playing with the API. Use swaggerfile mentioned above as a reference.

JWT based authorization

The application provides an API that protected by a very simple JWT autherization. For simplicity, the application use hard-coded user with the following credentials: username - user, password - password.

To authorize requests to the API, we have to get the token by calling the /signin endpoint with appropriate credentials. For example:

Please note: the output may differ

curl -X POST http://localhost:19000/signin --data '{"username":"user","password":"password"}'

{"token":"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VybmFtZSI6InVzZXIiLCJleHAiOjE2NTg4NTUwMTV9.JligE1ZNoARJ1Uf8IulyEBbQwE5QdHDdLh7gYScTCYw","expires":1658855015}

Then we can use this token to authorize requests to our API:

curl -X GET -H 'Authorization: eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VybmFtZSI6InVzZXIiLCJleHAiOjE2NTg4NTUwMTV9.JligE1ZNoARJ1Uf8IulyEBbQwE5QdHDdLh7gYScTCYw' http://localhost:19000/api/v1/todos

{"has_more":false,"total_count":0,"data":[]}

expires contains the timestamp that we can use to identify when the token will be expired. To renew the token, we can call /refresh and get back renewed token (we have 30 seconds window to renew it):

curl -X GET -H "Authorization: eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VybmFtZSI6InVzZXIiLCJleHAiOjE2NTg4NTUwMTV9.JligE1ZNoARJ1Uf8IulyEBbQwE5QdHDdLh7gYScTCYw' http://localhost:19000/refresh

{"token":"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VybmFtZSI6InVzZXIiLCJleHAiOjE2NTg4NTY5MzV9.6LCtnVntqmn2i5fbLakM6Y13T6HztKi2nfkc4-IUL2w", "expires":1658856935}

Deployment

The application delivered within staged Dockerfile. Thus we can easily use different strategies to deploy it.

Kubernetes

If we want to deploy the application using Kubernetes, we can simply build a docker image out of the Dockerfile and put it into your favorite registry. Application is stateless, so we can schedule it easily using Kubernetes Deployment object. We should create ConfigMap in order to provide configuration .yaml file It is meant that before starting application there should be Postgres DB running. We should keep this in mind and don't forget to provide the DSN to the application.

Application exposes 2 endpoints to check health/readiness: /healthz & /readyz

Run as a binary

If we want to deploy this app as a binary, we can do it using enormous possible ways. The repo provides Makefile. By running make build the binary for current OS/arch will be compiled. make build-release will produce linux-specific production ready binary, consult Makefile to learn more.

Test

To run test, simply type a command make test.

To improve

1. Test coverage
2. Provide k8s config files as a reference and a way to deploy
3. Extend API to include comments/labels into the TODOs response
4. Multi-user API, with authentication/authorization flow ...