Programming - Software Developer's Tour

Understanding K-Nearest-Neighbors (KNN) – Essential Machine Learning Algorithm

Paweł Majewski — Sun, 29 Jun 2025 01:21:34 +0000

A quick introduction to this algorithm

KNN is used to determine the class of our sample data. To understand this more clearly, please take a look at the plot below. We have two groups of samples: blue ones and red ones. We know a priori which samples are blue and which are red. Now, however, a new sample has arrived – a green one. Our task is to determine whether this sample better fits with the reds or the blues.

As you might guess, the simplest way to assign class A or B to a new sample is to calculate the distance between our sample and other points. Why does KNN have a ‘K‘ in its name? The ‘K‘ represents the number of nearest neighbors used to predict a classification.

How do we calculate distance between samples?

The easiest way to calculate the distance is to use a formula we all know from primary school, called the Euclidean distance. Of course, there are other algorithms that are better depending on the distribution of our data.

\sqrt{{(x ₂ - x ₁)}^{2} + {(y ₂ - y ₁)}^{2}}

Just code this function

def euclidean_distance(a, b):
  return np.sqrt(np.sum(a - b) ** 2)

Now, we need to find the K nearest neighbors. First, we need to calculate the distance between each point and the test point.

def get_neighbors(X, y, test_point, k):
    distances = []
    for i in range(len(X)):
        distance = euclidean_distance(X[i], test_point)
        distances.append((X[i], y[i], distance))
    distances.sort(key=lambda x: x[2])
    neighbors = distances[:k]
    return neighbors

In this step, we will calculate class for our test point.

This is the result of get_neighbors function.

As you can see, we have three instances of class 0 and two instances of class 1. Therefore, when the KNN algorithm is run on this data with K equal to 5, it will classify the test point as class 0, since it is nearest to this class.

This is how it looks like

Our test point has three connections with the blue ones and only two connections with the red ones.

d(P, Q) = \sqrt{\sum i_{1}^n ((p_i - q_i)^2)}

Our function for calculating the Euclidean distance, called euclidean_distance, is designed to solve real-life problems that are not limited to two dimensions but can also handle n-dimensional scenarios.

The post Understanding K-Nearest-Neighbors (KNN) – Essential Machine Learning Algorithm first appeared on Software Developer's Tour.

A Step-by-Step Guide to Web Scraping for Beginners

Paweł Majewski — Sat, 21 Jun 2025 08:38:00 +0000

In this article, I would like to raise the subject of web scraping.

Before we start, we have to consider web scraping on several layers: legal and ethical.

Let’s tackle the topic of the legality of web scraping first.

Fundamentally, the law depends on the country you live in. You should remember about:

Authors’ right

Terms of Service

General Data Protection Regulation (GDPR, European Union)

Computer Fraud and Abuse Act (CFAA, USA)

Outside the regulations you have to make sure that you are not affecting on target’s infrastructure by for example DDOSing target’s site.

Let’s get to practice

We want to get all titles from the blog section.

When we analyze the blog section, we can see that the title is an a tag inside a div with the class blog-page-title.

The steps we need to take to accomplish our task:

1. Go to the website.
2. Wait for the page to load until there is a div with the class “blog-page-title” and an a tag inside it.
3. Find the a tag inside the div with the class mentioned above.
4. Print the text of these a tags.

from selenium import webdriver
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC

url = "https://pawelmajewski.com/blog"
driver = webdriver.Chrome()
driver.get(url)
div_class = "blog-page-title"
wait = WebDriverWait(driver, 3)
wait.until(EC.presence_of_element_located((By.CSS_SELECTOR, f"div.{div_class} a")))
a_tags = driver.find_elements(By.CSS_SELECTOR, f"div.{div_class} a")

for a_tag in a_tags:
    print(a_tag.text)

driver.quit()

I recommend using the Selenium library both for writing your own web scrapers as well as for writing, for example, integration tests.

The post A Step-by-Step Guide to Web Scraping for Beginners first appeared on Software Developer's Tour.

Effortless API Testing with Visual Studio Code: Quick and Easy Guide

Paweł Majewski — Tue, 10 Jun 2025 20:15:10 +0000

Have you ever wondered how to test your API endpoints? There are many options to do so. You can use:

Swagger

Postman

Insomnia

(other less known software)

or our star for today…

Visual Studio Code

Visual Studio Code (VSC) is a really powerful tool. Calling VSC an IDE wouldn’t be entirely accurate, as its functionality largely depends on extensions.

One time VSC can be typical text editor, and the other time you can use VSC to manage your Linux server files over SSH using https://marketplace.visualstudio.com/items?itemName=ms-vscode-remote.remote-ssh.

Firstly, let’s create a new Web API project and modify the default WeatherForecastController with the code below.

using Microsoft.AspNetCore.Mvc;
using Microsoft.Net.Http.Headers;

namespace TestApi.Controllers;
[ApiController]
[Route("[controller]")]
public class WeatherForecastController : ControllerBase
{
    [HttpGet("get")]
    public IActionResult Get()
    {
        return Ok(Enumerable.Range(1, 5));
    }

    [HttpGet("get-protected")]
    public IActionResult GetProtected()
    {
        if (!Request.Headers.ContainsKey(HeaderNames.Authorization) || Request.Headers[HeaderNames.Authorization] != "Bearer testKey")
            return Unauthorized();

        return Ok(Enumerable.Range(1, 5));
    }

    [HttpPost("post")]
    public IActionResult Post([FromBody] TestRequest request)
    {
        return Ok(request);
    }

    public class TestRequest
    {
        public required string Name { get; set; }
        public int Age { get; set; }
    }
}

As you can see, there are three endpoints that we will test today:

Get

GetProtected

and Post

These three endpoints will cover all our base cases, so let’s get started!

Install REST Client extension

https://marketplace.visualstudio.com/items?itemName=humao.rest-client

Create file with .http extension

Naming convention It's a good practice to name your files with words that clearly describe their purpose. Today, I'll test the endpoints in WeatherForecastController, so I've named the file WeatherForecastControllerTests.http.

First sending of the request

The arrow points to the button which sends the request, and on the right panel, you can see the response from the server.

GET https://localhost:7081/WeatherForecast/get

Let's make authorized request

We’ve received a 401 Unauthorized error. I’ve created a fake authentication based on a Bearer token. If you are not familiar with Bearer tokens, read more https://docs.github.com/en/rest/authentication/authenticating-to-the-rest-api?apiVersion=2022-11-28

What should we do? Just add an Authorization header

GET https://localhost:7081/WeatherForecast/get-protected
Authorization: Bearer testKey

How about POST http method?

Nothing easier!

What is worth highlighting is that VSC automatically suggests possible HTTP MIME types.

The full request

POST https://localhost:7081/WeatherForecast/post
Content-Type: application/json

{
    "name": "pawelmajewski.com",
    "Age": 22
}

Improvements

We’ve got this. Our tests work perfectly, but… what if the base URL changes?

Variables

Yes!We will replace repetitive elements with variables.

@baseUrl = https://localhost:7081
@apiToken = Bearer testKey
@contentTypeApplicationJson = application/json

GET {{baseUrl}}/WeatherForecast/get

###

GET {{baseUrl}}/WeatherForecast/get-protected
Authorization: {{apiToken}}

###

POST {{baseUrl}}/WeatherForecast/post
Content-Type: {{contentTypeApplicationJson}}

{
    "name": "pawelmajewski.com",
    "Age": 22
}

Now, I’m happy with this code. As you can see, VSC is a very powerful tool with an enormous number of extensions that make VSC a multitool.

Additionally, what makes this tool very simple and fast to use for programmers is that here you simply write code, without needing to learn a new UI to do anything.

The post Effortless API Testing with Visual Studio Code: Quick and Easy Guide first appeared on Software Developer's Tour.

Simplify Your React Forms with React Hook Form and Yup

Paweł Majewski — Mon, 12 May 2025 09:52:00 +0000

Here we are! The first post about frontend technology has been written. Until now, only backend and DevOps topics have been covered.

Today I want to show you one of the top libraries that allows us to manage forms easily.

React Hook Form and Yup

What is React Hook Form?

It’s one of the most frequently used library to manage forms with React. As we can read on the project’s site

Performance is one of the primary reasons why this library was created.

https://react-hook-form.com/faqs

What is Yup?

Yup is a library used for creating schemas that help in parsing and validating values at runtime.

npm install react-hook-form
npm install @hookform/resolvers yup

I have created simple form using bootstrap.

In a standard case, we would create a useState or useRef for each field. It would result in a massive amount of code. However, it’s a good approach when you are new to React and still learning. By implementing it yourself, you will learn how React works and how to solve basic problems that surprise beginners. Okay, let’s get back to the main thread.

I have created React with typescript, so let’s create a type that defines our form.

type FormData = {
    firstName: string;
    lastName: string;
    age: number;
    email: string;
    password: string;
}

Then is good time to define our validation schema for this model.

const schema: ObjectSchema = yup
    .object({
        firstName: yup.string().required("FirstName cannot be empty"),
        lastName: yup.string().required("LastName cannot be empty"),
        age: yup.number().positive("Age must be positive number").integer().required("Age cannot be empty").typeError("Age must be a number"),
        email: yup.string().required("Email cannot be empty").matches(RegexConsts.email, "Email must contain @ and . characters"),
        password: yup.string().required("Password cannot be empty"),
    });

Yup provides many validation methods. I won’t waste your time, so here I will only touch on the matches method.

Matches method allows us to define regex pattern to validate this field. The second parameters means failed validation message.

export const RegexConsts = {
    email: /.*[@].*[.][a-zA-Z]{2,}/
}

React Hook Form validates forms based on configuration, in this case after submitting the form.

Custom validation for age

yup.number().positive("Age must be positive number").integer().required("Age cannot be empty").typeError("Age must be a number")

When validation is successfully completed, the submit method returns our form object with the filled fields.

The full file looks as presented below

import { useForm } from "react-hook-form"
import { yupResolver } from "@hookform/resolvers/yup"
import * as yup from "yup"
import { RegexConsts } from "../consts/regexConsts"
import { ObjectSchema } from "yup"

type FormData = {
    firstName: string;
    lastName: string;
    age: number;
    email: string;
    password: string;
}

const schema: ObjectSchema = yup
    .object({
        firstName: yup.string().required("FirstName cannot be empty"),
        lastName: yup.string().required("LastName cannot be empty"),
        age: yup.number().positive("Age must be positive number").integer().required("Age cannot be empty").typeError("Age must be a number"),
        email: yup.string().required("Email cannot be empty").matches(RegexConsts.email, "Email must contain @ and . characters"),
        password: yup.string().required("Password cannot be empty"),
    });


export default function RegisterForm() {
    const {
        register,
        handleSubmit,
        formState: { errors },
    } = useForm({
        resolver: yupResolver(schema),
    })

    const onSubmit = (data: FormData) => console.log(data)

    return (
        <>
            
                
                    
                        
                            First name
                                
                            
                            
                                {errors.firstName?.message}
                            
                        
                        
                            Last name
                                
                            
                            
                                {errors.lastName?.message}
                            
                        
                        
                            Age
                                
                            
                            
                                {errors.age?.message}
                            
                        
                        
                            Email
                                
                            
                            
                                {errors.email?.message}
                            
                        
                        
                            Password
                                
                            
                            
                                {errors.password?.message}
                            
                        
                    
                    
                        
                    
                
            
        
    )
}

I highly recommend using React Hook Form combined with Yup validation in your daily work with React.

The post Simplify Your React Forms with React Hook Form and Yup first appeared on Software Developer's Tour.

A Comprehensive Guide to Preventing SQL Injection

Paweł Majewski — Sun, 04 May 2025 10:30:14 +0000

I recently wrote an article which is the first part about SQL Injection SQL Injection in C# Application: An Example.

Today, we will continue this topic answering the question asked in the previous post “How can we avoid SQL Injection?“

Entity Framework

The first option is to use Entity Framework as ORM in our application. Entity Framework uses LINQ-to-Entities parametrized queries preventing SQL Injection attacks.

I’m using here the same database form previous post created with Scaffolding (Reverse Engineering)

dotnet ef dbcontext scaffold "Data Source=localhost;Initial Catalog=TestSqlInjection;Trusted_Connection=yes;Integrated Security=true;TrustServerCertificate=True;" Microsoft.EntityFrameworkCore.SqlServer

As you can see, we have 3 users in database. Let’s try to select user wtih email “test@example.com“.

using Sql_Injection;

var context = new TestSqlInjectionContext();
var email = "test@example.com";
var password = "qwe";
var user = context.Users
    .Where(x => x.Email == email && x.Password == password)
    .FirstOrDefault();

Console.WriteLine($"User: {user?.Email}");

The result of this code will be

“User: test@example.com”

Let’s use it and try SQL Injection on this query

using Sql_Injection;

var context = new TestSqlInjectionContext();
var email = "test@example.com";
var password = "' or 1 =1--";
var userSqlInjection = context.Users
    .Where(x => x.Email == email && x.Password == password)
    .FirstOrDefault();

Console.WriteLine($"SqlInjectionUser: {userSqlInjection?.Email}");

The result:

SqlInjectionUser:

It’s because userSqlInjection has not been found. SQL injection won’t work when we use an ORM like Entity Framework because the inputs are automatically sanitized. Furthermore Entity Framework restricts direct SQL execution, developers are advised to use parametrized queries or LINQ in order to meet the security requirements.

Is this the only way to prevent SQL injection? Of course not, more methods will be presented on the blog soon.

The post A Comprehensive Guide to Preventing SQL Injection first appeared on Software Developer's Tour.

Unlocking the Power of LINQ: 7 Rarely Used Queries You Should Know

Paweł Majewski — Wed, 16 Apr 2025 15:47:08 +0000

Today, I would like to talk about LINQ. LINQ queries are ubiquitous in the C# world, whether you’re working with WebAPI, MVC, or Windows Forms applications.

But honestly, we are using only a small part of the LINQ methods.

Where, Select, OrderBy, ThenBy / ThenByDescending, GroupBy, Join, First, FirstOrDefault, Single, SingleOrDefault, Count, Any, All and some aggregate functions for sequence of numeric values like Sum, Min, Max, Average.

Rarely used queries

Let’s look on less known methods.

1. Distinct

Distinct exactly like in SQL selects unique elements

int[] array = [1, 2, 3, 1, 2];

foreach(int i in array.Distinct())
{
    Console.Write($"{i}, ");
}

And the output is: 1, 2, 3,

2. Intersect

Intersect returns elements existing in both collections

int[] array1 = [1, 2, 3, 4];
int[] array2 = [1, 2];

foreach (int i in array1.Intersect(array2))
{
    Console.Write($"{i}, ");
}

And the output is: 1, 2,

3. Except

Except returns elements from first collection, not existing in the second one

int[] array1 = [1, 2, 3, 4];
int[] array2 = [1, 2];

foreach (int i in array1.Except(array2))
{
    Console.Write($"{i}, ");
}

And the output is: 3, 4,

4. SkipWhile

SkipWhile checks the condition for each element in the sequence. When element does not match this predicate, filtering stops, and the rest of the sequence is returned.

int[] array1 = [1, 2, 3, 4, 1, 2];

foreach (int i in array1.SkipWhile(x => x < 3))
{
    Console.Write($"{i}, ");
}

And the output is: 3, 4, 1, 2,

but

int[] array1 = [5, 1, 2, 3, 4, 1, 2];

foreach (int i in array1.SkipWhile(x => x < 3))
{
    Console.Write($"{i}, ");
}

And the output is: 5, 1, 2, 3, 4, 1, 2,

Because 5 doesn’t match condition “x < 3” and filtering has stopped.

5. Zip

Zip Combines two or more sequences into a single one, output collection length will be the smallest input collection.

int[] array1 = [5, 1, 2, 3, 4, 1, 2];
int[] array2 = [0, 1, 2];

foreach (var i in array1.Zip(array2, (x, y) => $"[{x}] = {y}"))
{
    Console.Write($"{i}, ");
}

And the output is: [5] = 0, [1] = 1, [2] = 2,

6. Prepend

Prepend is similar to Append, but it adds an element at the beginning.

int[] array1 = [1, 2];

foreach (var i in array1.Prepend(3))
{
    Console.Write($"{i}, ");
}

And the output is: 3, 1, 2,

7. ToLookup

ToLookup is one of the most interesting LINQ method for me. It allows us to create groups just like GroupBy, but with immediate execution.

Car[] array1 = [new Car("Name1", "AwesomeBrand"), new Car("Name2", "AwesomeBrand"), new Car("Name3", "AwesomeBrand"), new Car("Name4", "NormalBrand")];

foreach (var i in array1.ToLookup(x => x.Brand))
{
    Console.WriteLine($"{i.Key}: {string.Join(',', i.Select(x => x.Name))}");
}

record Car(string Name, string Brand);

And the output is:

AwesomeBrand: Name1,Name2,Name3
NormalBrand: Name4

I wanted highlight these seven useful yet often forgotten methods in LINQ, even though there are many more worth exploring.

The post Unlocking the Power of LINQ: 7 Rarely Used Queries You Should Know first appeared on Software Developer's Tour.

Chat GPT and refactor?

Paweł Majewski — Tue, 08 Apr 2025 19:56:00 +0000

In this post How to generate website using Chat GPT I checked how ChatGPT is doing with generating the layout.

Today, we will see how it copes with the refactoring.

Variables chain

var x = 5;
var a = x;
var b = a;
var c = b;
Console.WriteLine(c);
var d = c;
var e = d;
var f = e + 1;
var g = f;
var h = g;
Console.WriteLine(g);

That’s not exactly what I meant.

Let’s try it differently.

Much better.

Complicated LINQ

int[] numbers = { 1, 2, 3, 4, 5 };

var result = numbers
    .Where(n => n % 2 == 0)
    .Select(n => n * n)
    .OrderByDescending(n => n)
    .Skip(1)
    .TakeWhile((n, index) => index < 2 || n > 10)
    .Sum();

Console.WriteLine(result);

Okay, time to ask ChatGPT what can we do with this piece of code.

I’m pasting here the code provided by ChatGPT, which allows you to verify that I’m not lying.

int[] numbers = { 1, 2, 3, 4, 5 };

var result = numbers
    .Where(n => n % 2 == 0)
    .Select(n => n * n)
    .OrderByDescending(n => n)
    .TakeWhile((n, index) => index < 2 || (index == 2 && n > 10))
    .Sum();

Console.WriteLine(result);

This code changes the result. The original code returned 4, but now it’s 20. Tell that to ChatGPT and watch how it will react.

int[] numbers = { 1, 2, 3, 4, 5 };

var result = numbers
    .Where(n => n % 2 == 0)
    .Select(n => n * n)
    .OrderByDescending(n => n)
    .TakeWhile((n, index) => index < 2 || (index == 2 && n > 10))
    .Sum();

Console.WriteLine(result);

Ok, let’s see.

ChatGPT said that this code would now return 4 as a result, but as you can see, it’s still 20.

My answer for this is

Yes, I know ChatGPT is capable of performing good refactoring, but we still need to double-check everything based on ChatGPT’s answer.

ChatGPT is wonderful tool, but again – it’s not medicine for every problem.

The post Chat GPT and refactor? first appeared on Software Developer's Tour.

Using Visual Studio Code to create .NET application

Paweł Majewski — Wed, 02 Apr 2025 17:21:33 +0000

Many of us use Visual Studio Code for frontend purposes, but how about using it to develop an .NET Applications?

I always use Visual Studio or Rider to develop .NET apps, but today, I’ll create a simple TODO application using Visual Studio Code to check if I’m able to create the application in a convenient way.

As first glance, we can see the “Create .NET Project” button which allows us to create an application with a few clicks.

I’ll choose standard MVC application as easiest way to test VSC in real application.

public class TodoItem
{
    public string Title { get; set; }
    public string Description { get; set; }
    public TodoItemStatus Status { get; set; }
}

I’ve modified HomeController and added TodoItems storage as static list

    public static List TodoItems = [
        new() {
            Title = "Write new post",
            Description = "Go to the pawelmajewski.com and add a new post",
            Status = TodoItemStatus.New
        }
    ];

And the result

Ok, let’s stop at this moment. I’ll write something about using Visual Studio Code and the feeling of using it instead of a dedicated IDE created for .NET applications.

As you can see, VSC doesn’t add namespace by default.

None of these options generate correct namespace automatically. The first option ends word to full “namespace” and the second one creates new namespace scope. Too bad I couldn’t find such an option to generate correct namespace with consideration for the folder structure.

I've found solution There are several extensions allowing create class with namespace automatically. You can read more about it Stack overflow

Summary

Visual Studio Code is amazing software with a vast amount of extensions. We can easily develop .NET applications using Visual Studio Code. There are many extensions available that make our job faster and more convenient. However, in my opinion, the usage of dedicated IDEs is a better option. I’ll stay with Rider / Visual Studio for .NET and Visual Studio Code for frontend 🙂

The post Using Visual Studio Code to create .NET application first appeared on Software Developer's Tour.

SQL Injection in C# Application: An Example

Paweł Majewski — Sat, 22 Mar 2025 19:18:38 +0000

Today, I'll show you some code vulnerable to SQL Injection. We will talk about good and bad practices. At the end I'll tell you which option is in my opinion the best one and why.

So... Let's get started...

Remember! Do not do anything that has been shown here in real applications; you may face legal and financial consequences.

To try to do something with SQL, we need a new table in our fake database.

Now, have a look at this ugly code…

What do we have here?

We have a POST endpoint /login which receives two strings, email and password.

This endpoint connects to the database and executes a command to select the first match of a user where the email and password (in plain text for testing purposes 🙃) match the provided data.

app.MapPost("/login", (string email, string password) =>
{
    const string connectionString = "Data Source=localhost;Initial Catalog=TestSqlInjection;Integrated Security=true;TrustServerCertificate=true;";
    using var connection = new SqlConnection(connectionString);
    using var command = connection.CreateCommand();
    command.CommandText = $"SELECT * FROM USERS WHERE Email = '{email}' AND Password = '{password}';";
    connection.Open();
    using var reader = command.ExecuteReader();
    reader.Read();
    
    return new User(Convert.ToInt32(reader["Id"]), reader["Email"].ToString()!);
})
.WithName("Login")
.WithOpenApi();

This code has couple of minor errors.

Command is concatenated from user’s data without any validation or using sql parameters.

Okay, let’s add some users

SELECT TOP (1000) [Id]
      ,[Email]
      ,[Password]
  FROM [TestSqlInjection].[dbo].[Users]

okay, there we have it. Let’s try to log in to our application with email “tester@example.com” and password “qwe“.

It looks like it’s working, doesn’t it? Don’t fall for it.

There are many things that can be done with our dirty backend code. Try logging in as “tester@example.com“ without knowing its password.

Let’s enter the following code as the password.

' or 1=1--

Explanation:

SELECT * FROM USERS WHERE Email = ‘{email}’ AND Password = ‘{password}‘;

The bolded part will be replaced with the password provided by the user. So, we are using a ‘ (single quote) as the end of the string. Do you see it? In the SQL clause, we have:

AND Password = ‘{password}‘;

When we use a single quote as the first character, we end this clause. Then we start a new clause, OR 1 = 1, which is always true.

What is the role of the — (SQL comment syntax) at the end?

Without this, our query would end with a single quote character, which is invalid SQL. We have to comment out the rest of the code after the password to ensure that our query will end with OR 1 = 1. Without this, we would get the following error:

Unclosed quotation mark after the character string ‘;
‘.

So the final clause is

SELECT * FROM USERS WHERE Email = 'tester@example.com' AND Password = '' or 1=1--';

Without security measures, someone could steal sensitive data. We can log in as any registered user.

If what we just did happened in a real application, it would definitely be a breach of GDPR in Europe.

Let’s get back on topic, though. Is that all we can do?

How about destroying data?

Of course system hacker can use that trick to destroy our data. Have a look:

'; update users set password = 'newpass'--

The final clause is

SELECT * FROM USERS WHERE Email = 'tester@example.com' AND Password = ''; update users set password = 'newpass'--';

And the result? Please

Our application has a bug. When no user has been found, we’re getting an exception, but that doesn’t matter because I wanted this application to have nothing but the essentials for testing our SQL.

Okay, what’s stopping us from clearing our table? Nothing.

Let’s check the following password

'; delete from users--

The final clause

SELECT * FROM USERS WHERE Email = 'tester@example.com' AND Password = ''; delete from users--';

Of course there’s an exception in our application but what’s happened in our database?

We lost all of them.

How can we avoid SQL Injection? That will be in the next post.

The post SQL Injection in C# Application: An Example first appeared on Software Developer's Tour.

How to create simple healthcheck in 5 minutes

Paweł Majewski — Tue, 04 Mar 2025 17:07:00 +0000

In the past, when I used a different hosting provider than I do now, I often encountered situations where the website was unavailable.

How can we check if a website is accessible without constantly refreshing it in the browser?

Just write your own app!

First, let’s think about what we want to do.

Application requirements

We want the ability to pass multiple URLs to the application
We haven't time, so our choice will be Console application😊
The application should inform us whether the request has returned a success status code or has failed.
To avoid being banned for DDoS, the application should incorporate a delay.between requests

You can read more about DDoS attack here.

Possibility of being banned You have to handle the given code very carefully, as placing the delay incorrectly could result in your application unintentionally launching DDoS attacks on the provided websites.

I will paste here entire code and then we’ll talk about it closer.

class Program
{
    static async Task Main(string[] args)
    {
        const string fileName = "sites.txt";
        TimeSpan delay = TimeSpan.FromSeconds(10);
        var fileContent = await ValidateFileAsync(fileName);
        var urlsWithClients = PrepareUrls(fileContent);

        await RunInfinity(urlsWithClients, delay);
    }

    static async Task IsAliveAsync(HttpClient httpClient)
    {
        try
        {
            var result = await httpClient.GetAsync(string.Empty);

            return result.IsSuccessStatusCode;
        } catch
        {

        }

        return false;
    }

    static async Task ValidateFileAsync(string filePath)
    {
        var doesFileExist = File.Exists(filePath);

        if (!doesFileExist)
        {
            Console.ForegroundColor = ConsoleColor.Red;
            Console.WriteLine("Couldn't find file");
            Console.ResetColor();

            return [];
        }

        var fileContent = await File.ReadAllLinesAsync(filePath);

        if(!fileContent.Any())
        {
            Console.ForegroundColor = ConsoleColor.Red;
            Console.WriteLine("There is no url in file");
            Console.ResetColor();
        }

        return fileContent;
    }

    static (string Url, HttpClient httpClient)[] PrepareUrls(IEnumerable urls)
    {
        return urls.Select(x => {
            var httpClient = new HttpClient
            {
                BaseAddress = new Uri(x)
            };

            return (x, httpClient);
        }).ToArray();
    }

    static Task RunInfinity(IEnumerable<(string Url, HttpClient httpClient)> urls, TimeSpan delay)
    {
        return Task.Run(async () =>
        {
            while (true)
            {
                foreach (var (x, httpClient) in urls)
                {
                    var isAlive = await IsAliveAsync(httpClient);

                    Console.Write($"{DateTime.Now:dd-MM HH:mm:ss}\tUrl: {x}\t");
                    if (isAlive)
                    {
                        Console.ForegroundColor = ConsoleColor.Green;
                        Console.WriteLine("OK");
                    }
                    else
                    {
                        Console.ForegroundColor = ConsoleColor.Red;
                        Console.WriteLine("FAILED");
                    }
                    Console.ResetColor();

                }

                await Task.Delay(delay);
            }
        });
    }
}

Let’s start analyzing 🙂

The main function

        const string fileName = "sites.txt";
        TimeSpan delay = TimeSpan.FromSeconds(10);
        var fileContent = await ValidateFileAsync(fileName);
        var urlsWithClients = PrepareUrls(fileContent);

        await RunInfinity(urlsWithClients, delay);

First two lines are our configuration. When creating this application, I assumed that the user will provide urls to check in file named sites.txt and this file will be next to .exe file. I aimed to keep it as straightforward as I could.

Then we have simple validation, I think we can just skip it. I added it only because I got an exception during development 😁

I didn’t want to read the file twice, so the validation function immediately passes on the read data.

    static (string Url, HttpClient httpClient)[] PrepareUrls(IEnumerable urls)
    {
        return urls.Select(x => {
            var httpClient = new HttpClient
            {
                BaseAddress = new Uri(x)
            };

            return (x, httpClient);
        }).ToArray();
    }

I think this is a good place to refactor. We don’t need a tuple with a string and HttpClient here. Since HttpClient already contains BaseAddress itself, we could retrieve this information directly from it.

Function IsAliveAsync is only wrapper for GetAsync method.

And last our system’s core, our heart beating method

        return Task.Run(async () =>
        {
            while (true)
            {
                foreach (var (x, httpClient) in urls)
                {
                    var isAlive = await IsAliveAsync(httpClient);

                    Console.Write($"{DateTime.Now:dd-MM HH:mm:ss}\tUrl: {x}\t");
                    if (isAlive)
                    {
                        Console.ForegroundColor = ConsoleColor.Green;
                        Console.WriteLine("OK");
                    }
                    else
                    {
                        Console.ForegroundColor = ConsoleColor.Red;
                        Console.WriteLine("FAILED");
                    }
                    Console.ResetColor();

                }

                await Task.Delay(delay);
            }
        });

What are we doing here? There are just two loops, one while loop, which with Task.Delay (line 24) will be our beating heart.

Inside while loop is loop over our urls and and that’s it, no philosophy involved.

Additionally, one “if” statement to check whether the request was successful or not.

How it works?

Create file with urls

Start application

At the end

I strongly encourage you to write simple applications. Writing one can take as much time as finding the right one online.

The post How to create simple healthcheck in 5 minutes first appeared on Software Developer's Tour.