Weather Forecast Service (WEATHY)

General

Prerequisites

You will need the following tools and SDKs to build, run and test the service:

• Microsoft Visual Studio 2019 (Windows)
• Microsoft .NET SDK 5.0 or later
• Docker (including docker-compose)
• WSL (Windows Subsystem for Linux) or a virtual machine with Linux

You may also need:

• Postman
• Heroku CLI

Build, run, test

How to build locally

You can build and run the service in Microsoft Visual Studio with the solution file.

Alternatively, you can build with Microsoft .NET SDK. To build the release version of the service, run:

dotnet build src/weathy.csproj -c release

In order to build the service with all dependencies, run:

dotnet publish src/weathy.csproj -c release

How to run tests

To run unit and integration tests, run:

dotnet test test/weathy-test.csproj

How to run locally

You can run the service directly from Visual Studio using the solution file.

Alternatively, you can run the service from the command line as follows:

dotnet run -p src/weathy.csproj -c release

How to build a docker container

To build a docker container with the service, run:

docker build -t weathy:latest src

How to run locally with docker-compose

To get the service running locally as a container, run:

docker-compose up -d

To stop the service, run:

docker-compose down

How to publish and run in Heroku

The following instructions show how to deploy and test the service in Heroku.

NOTE: You will need to create your own Heroku app and use the app name in the commands below.

Login to Heroku:

heroku login

Login to Heroku Container service:

heroku container:login

Build the Dockerfile in the current directory and tag it as follows:

docker tag weathy registry.heroku.com/<your-heroku-app-name>/web

Push the Docker image to Heroku.

docker push registry.heroku.com/<your-heroku-app-name>/web

Release the newly pushed images to deploy the service.

heroku container:release web --app <your-heroku-app-name>

Configuration

The service can read configuration settings from:

• appsettings.json file
• environment variables
• command line arguments

The above order is also the order of preference, which means that settings could be overridden in that order. The later value will be used.

appsettings.json

This is the basic settings file for general settings that could be applied in all environments. These settings could be overridden by appsettings.{environment}.json if there are settings specific to the environment. {environment} defines the execution environment, for example: development, production, test.

Environment variables

The environment variables shall be created with WEATHY_ prefix. When added to the configuration the prefix is removed. For example:

on Windows:

set WEATHY_ENABLE_SWAGGER_UI=true

on Linux:

export WEATHY_ENABLE_SWAGGER_UI=true

Command line arguments

Command line arguments could be defined like this when running the program:

dotnet Weathy.dll ENABLE_SWAGGER_UI=true

Configuration settings

Below are the configuration settings supported by the service:

Name	Description	Default
ASPNETCORE_URLS	Defines URL(s) that the service is listening on. The setting shall be either defined as the environment variable or --urls command-line argument.	http://localhost:5000
ASPNETCORE_ENVIRONMENT	Defines the hosting environment. The service loads configuration settings based on the environment name, i.e. appsettings.json and appsettings.{Environment}.json	Empty
ENABLE_SWAGGER_UI	Enables Swagger UI at /swagger endpoint	false
WEATHER_API_KEY	API key to use Weather API	Empty
WEATHER_API_BASEURL	The base URL of Weather API endpoint	https://api.weatherapi.com/
WEATHER_API_NUMDAYS	The number of days of forecast required	5

REST API

The service implements the REST API with the following endpoints. You can run the service with ENABLE_SWAGGER_UI enabled and find the API documentation at http://localhost:5000/docs.

GET /api/v1/weather

Retrieve "random" weather forecast

Example:

curl -X 'GET' 'http://localhost:5000/api/v1/weather'

Response:

[
  {
    "date": "2021-05-07T19:40:04.7780682+03:00",
    "temperatureC": -10,
    "temperatureF": 15,
    "summary": "Hot"
  },
  {
    "date": "2021-05-08T19:40:04.7783563+03:00",
    "temperatureC": 20,
    "temperatureF": 67,
    "summary": "Mild"
  },
  {
    "date": "2021-05-09T19:40:04.77836+03:00",
    "temperatureC": 39,
    "temperatureF": 102,
    "summary": "Bracing"
  },
  {
    "date": "2021-05-10T19:40:04.7783603+03:00",
    "temperatureC": -4,
    "temperatureF": 25,
    "summary": "Cool"
  },
  {
    "date": "2021-05-11T19:40:04.7783606+03:00",
    "temperatureC": 3,
    "temperatureF": 37,
    "summary": "Cool"
  }
]

GET ​/api​/v2​/weather?city=<city_name>

Retrieve actual weather forecast for the requested city.

Example:

curl -X 'GET' 'http://localhost:5000/api/v2/weather?city=london'

Response:

[
  {
    "date": "2021-05-06T00:00:00",
    "temperatureC": 5,
    "temperatureF": 40,
    "summary": "Patchy rain possible"
  },
  {
    "date": "2021-05-07T00:00:00",
    "temperatureC": 9,
    "temperatureF": 48,
    "summary": "Partly cloudy"
  },
  {
    "date": "2021-05-08T00:00:00",
    "temperatureC": 8,
    "temperatureF": 46,
    "summary": "Moderate rain"
  }
]

Logging

The service employs Serilog library instead of the standard logging middleware in Microsoft .NET. Log events are printed to the console and recorded to the log file. Default logging level and settings can be defined/changed in the Serilog section of appsettings.json file.

The following example shows how to enable different log levels for Microsoft and system components:

  "Serilog": {
    "MinimumLevel": {
      "Default": "Debug",
      "Override": {
        "Microsoft": "Debug",
        "System.Net.Http.HttpClient": "Error"
      }
    },