RESTful Weather application

Tech Stack:

• Java 11
• SpringBoot
• Lombok
• Spring Data JPA
• H2 in-memory DB
• Mockito
• MockMvc
• Wiremock
• Flyway

Topics

1. How to run this application
2. How to access the spring boot restful application
3. Design and implementation
4. Assumptions

How to run this application

• Navigate the root folder /weather-app under the command line
• Run the command to build the whole project: gradle clean build
• Run the command to start the application: java -jar ./build/libs/weather-app-0.0.1-SNAPSHOT.jar --openweather.api.token={your openweather api token}
• If you want to run in IDE, please check application.yml to change "FIXME" in token part

How to access the spring boot restful application

you can access via: http://localhost:8090/weather?city={city}&country={two characters country code}&token={can be found in data init}

This is the example: http://localhost:8090/weather?city=shanghai&country=cn&token=ba3a5f28-eddf-43ee-8202-88682f23858e

Design and implementation

• Implement a separate RateLimitService for rate limitation, using a simple way like sliding window to count the requests for last hour. For calculating, stored requests history as token usages into a table.
• Hibernate builds the entity layer
• Flyway prepares the initial data
• Spring Data JPA builds the repository layer
• Mockito and MockMvc unit test service and controller layer, Wiremock for 3rd party API tests
• Defined customised exceptions for better error handling and all the exceptions can be centrally handled in one place (ControllerExceptionHandler.java)
• Lombok automatically generates getter,setter, constructor, hashcode, log etc.
• High test coverage

Assumptions

• didn't consider rate limitation in Openweather API side
• since the overall traffic is pretty low, didn't choose 3rd party lib for rate limits and faster ways for storing requests history