Weather Forecasting Rest API version 2.0

Getting started

This application is build on spring boot v 2.4.2.Release To use Weather Forecasting Rest API, User need to register with http://openweathermap.org service to get APIKEY, which will be added in application.properties file.

Prerequisite

• Maven 3.5.2
• Eclipse Oxygen
• Java 8

Build

This application uses MAVEN 3.5.2 for build. Use below command at the root directory of this project

mvn clean install

Run

Once the build is successful, you can find the jar file of this application in ~/target folder Run the jar file using below command

java -jar .\target\weatherforecast-0.0.1-SNAPSHOT.jar

Access API

http://localhost:9090/data/{cityname} e.g. http://localhost:9090/data/berlin

response Example:

[{"date":[2019,1,14],"avg_day_temp":34.705,"avg_night_temp":35.3625,"avg_pressure":1005.36375},{"date":[2019,1,15],"avg_day_temp":35.385,"avg_night_temp":36.5425,"avg_pressure":1016.5550000000001},{"date":[2019,1,16],"avg_day_temp":42.81999999999999,"avg_night_temp":43.102500000000006,"avg_pressure":1010.2175}]

Access Swagger UI

http://localhost:9090/swagger-ui.html This can be used for manual testing

Unit Testing

To Test Application :

you can run com.weatherforecast.WeatherForcastApplicationTests

Integration Testing

To Test Integration with openweathermap.org:

you can run com.weatherforecast.integration.WeatherForecastIntegrationTest

Reasoning and Motivations

At first, I created a high level design:

API call-->/data/{city}-->dataservice-->openweathermap forecast API-->dataservice-->response for /data

Note: In case study, I assume that, Average is referring to mean and not to mode or median

Following are my reasoning and motivations through out this case study:

• When ever I develop an API, I keep in mind the performance, maintainability, easy to test, and easy to read.
• code should be well commented for easy understanding.
• Bootstrap the application by creating project at https://start.spring.io/.
• Imported to eclipse for development, my preferable IDE.
• All the variables, which are subject to change were added in properties file, this helps in easy maintenance.
• Java 8 features like Lambda expressions with Stream API is used for better understanding and less code.
• Collections like ArrayList is used to collect data for different days.
• POJO classes are created to get the required forecast data from openweathermap.org.
• POJO class is created for /data response.
• Exceptions are created to throw the correct error code according to the REST API conventions.
• Input is getting validated at the starting of API call and appropriate error code is thrown.
• Caching is enabled in this application to cached the forecast data from openweatherapi.org, to improve the performance for multiple same API calls.
• server port is also changed to 9090 because mainly 8080 is already taken in several machines.
• I have used multiple functions to divide the logic and increase the code reuse.
• LocalDate and LocalTime is added in the response of openweathermap.org, which makes it easy to filter the entries according to the date and time.