Wadhwani AI Assignment

---

Overview

This microservice is system for adding, managing and getting information regarding farmers. The key endpoints include:

• Upload a csv file containing details of users
• Get details of a specific user given their id (autogenerated)
• Get details of the user in a specific language (English, Hindi, Marathi, Telugu, Punjabi)
• Register an auth user
• Login as an auth user to use endpoints.

Components

• Clients: mobile app or web app
• User Service microservice: To manage all functionalities regarding user management
• MySQL: SQL DB to store details on users and auth users
• Redis: used for caching API requests
• Cloud Translate: translation APIs for fetching details in specific language.

Flows

• Register:
  • accepts username and password and inserts them to db.
  • password in encrypted before storing.
• Login:
  • accepts username and password and checks them against the entry in db.
  • generates jwt token if successful.
  • jwt token is used for authorization and authentication of other endpoints.
• Upload User:
  • accepts a file upload (multipart upload) in the request of type csv.
  • parses csv to get users details.
  • stores users details in db.
• Get User:
  • accepts user id in url and language in headers
  • first tries to fetch details from cache.
  • if cache misses, we fetch details from db and hit translation service to get translated details.
  • if cache hits, we send the response as is.

Assumptions and Considerations

• Caching is used to reduce the number of translation api hits (reduces dependency and cost-efficient)
• All repositories and services have interfaces implemented for easy migration to other dbs/ third party services.
• Passwords are encrypted before storing for data security. Matching is done similarly.
• Upload user and Get user are authenticated APIs, you need a jwt generated via login to access them.

Setup

# Install golang
# Install mysql (setup env)
# Install redis (setup env)
# Add google credential files (setup env)

# Install all dependencies in go.mod
go mod download

# Build project
go build user-service

# Run project
./user-service

# Install go-swagger
brew tap go-swagger/go-swagger
brew install go-swagger

# Serve swagger doc (API Documentation)
swagger serve -F=swagger ./docs/swagger.json

# Get coverage
go tool cover -func profile.cov

Testing

Find the user service collection yaml file in assets/user-service-endpoints.yml

Improvements

• File Upload can use a file storage system (S3)
  • Separate storage for files
  • pipeline to parse files and send messages to the service to add/update data
• Parallelism
  • Use goroutines and sync groups to parallelize some functionalities (dto, csv parsing)