Category: .NET Core

EventId in logging

Finally I have realized why it is necessary to have EventId logged and it is not the same life without him – especially in the microservices context.

So why it is not enough

public static void LogError (this Microsoft.Extensions.Logging.ILogger logger, Exception? exception, string? message, params object?[] args);

and I strongly recommend

public static void LogError (this Microsoft.Extensions.Logging.ILogger logger, Microsoft.Extensions.Logging.EventId eventId, string? message, params object?[] args);

Let’s assume that you have something like a log aggregator , such as ElasticSearch

You want to see from the error message , at a glace the details of the problem – like the root cause , some additional data – without going into details of the object. Something like “Drive C: too full to write” or “Drive D: too full to write” .

Also, you want to see how many error of this kind occurs – how many “ too full to write “ have occured ?

You ca put the message like “ Drive too full to write – look into details ” – and count those – but the investigators will have to click on more time to find what drive it is.

Welcome to EventID – https://docs.microsoft.com/en-us/dotnet/api/microsoft.extensions.logging.eventid?view=dotnet-plat-ext-6.0 .

You can have a

var evt = new EventId(7000,”DriveTooFull” );

logger.LogError(evt, “Drive C: too full to write” )

Now you have the message detaisl and you can count how many 7000 ( or DriveTooFull ) errors you have

[NuGet]: HealthCheck

This is a Nuget that I use in all projects when I am making any ASP.NET Core prohect

Link: https://www.nuget.org/packages/AspNetCore.HealthChecks.UI/

Site: https://github.com/Xabaril/AspNetCore.Diagnostics.HealthChecks

What it does: Implements health checks for your site and/or dependencies

Usage:

public void ConfigureServices(IServiceCollection services)
{

//code
services.AddHealthChecks()
.AddSqlServer(myConnectionString)
}

[NuGet]: Polly

This is a Nuget that I use in all projects when I am making HTTP calls..

Link: https://www.nuget.org/packages/polly

Site: https://github.com/App-vNext/Polly

What it does: Implements all kind of policies for retrying – see https://github.com/App-vNext/Polly/wiki/Transient-fault-handling-and-proactive-resilience-engineering

Usage: – copied from https://docs.microsoft.com/en-us/dotnet/architecture/microservices/implement-resilient-applications/implement-http-call-retries-exponential-backoff-polly

static IAsyncPolicy<HttpResponseMessage> GetRetryPolicy()
{
     return HttpPolicyExtensions
         .HandleTransientHttpError()
         .OrResult(msg => msg.StatusCode == System.Net.HttpStatusCode.NotFound)
         .WaitAndRetryAsync(6, retryAttempt => TimeSpan.FromSeconds(Math.Pow(2,
                                                                     retryAttempt)));
}

Dependent Framework Versioning

There are multiple ways to version a software . I have used SemanticVersioning ( https://semver.org/ ) and Calendar Versioning ( https://calver.org/ )

Of course , there are others – please read https://en.wikipedia.org/wiki/Software_versioning – interesting versioning based on e or PI .

However , I want to propose a new standard :

Dependent Framework Versioning

The major version of a package is the same of the major version of the framework that is installed on . For the others , can be calver or semver or any other – depending on your choice.

For example , I have made a BlocklyAutomation package for .NET Core 3 – and it is versioned like

For .NET Core 3 – https://www.nuget.org/packages/NetCore2Blockly/3.2022.224.16

For .NET Core 5– https://www.nuget.org/packages/NetCore2Blockly/5.2022.210.2007

Why is this ? To be easy identified by the users of the package. If I have one user that have an app on .NET Core 3 and other on .NET Core 5, how can they identify easy what is the latest version for this package ? With this approach , this can be done just looking on the major version corresponding with the framework version (and yes, I use calver versioning for the rest – yyyy.Md.Hm)

[NuGet]:Scriban

This is a Nuget that I use in almost every .NET Core project to get the error details ( 500)

Link: https://www.nuget.org/packages/Scriban/

Site: https://github.com/scriban/scriban

What it does: Like Razor – a template intepreter

Usage:

var template = Template.Parse(@"
<ul id='products'>
  {{ for product in products }}
    <li>
      <h2>{{ product.name }}</h2>
           Price: {{ product.price }}
           {{ product.description | string.truncate 15 }}
    </li>
  {{ end }}
</ul>
");
var result = template.Render(new { Products = this.ProductList });

Using files in a ASP.NET Core nuget package

To use static (html) files in a NUGET ASP.NET Core package to be displayed I have used the following solution

1. Put the files in a folder ( in my case , blocklyAutomation )

2. Put in the .csproj to embed the files

<ItemGroup>
   <EmbeddedResource Include=”BlocklyAutomation\**\*”>
     <CopyToOutputDirectory>Never</CopyToOutputDirectory>
   </EmbeddedResource>
</ItemGroup>

3. Create an extension to use it

;

public static void UseBlocklyUI(this IApplicationBuilder appBuilder)
{
    var manifestEmbeddedProvider =
            new ManifestEmbeddedFileProvider(Assembly.GetExecutingAssembly());
    var service = appBuilder.ApplicationServices;
    mapFile("blocklyAutomation", manifestEmbeddedProvider, appBuilder);


}

private static void mapFile(string dirName, IFileProvider provider, IApplicationBuilder appBuilder)
{
    var folder = provider.GetDirectoryContents(dirName);
    foreach (var item in folder)
    {
        if (item.IsDirectory)
        {
            mapFile(dirName + "/" + item.Name, provider, appBuilder);
            continue;
        }
        string map = (dirName + "/" + item.Name).Substring(dirName.Length);
        appBuilder.Map(map, app =>
        {
            var f = item;

            app.Run(async cnt =>
            {
                //TODO: find from extension
                //cnt.Response.ContentType = "text/html";
                using var stream = new MemoryStream();
                using var cs = f.CreateReadStream();
                byte[] buffer = new byte[2048]; // read in chunks of 2KB
                int bytesRead;
                while ((bytesRead = cs.Read(buffer, 0, buffer.Length)) > 0)
                {
                    stream.Write(buffer, 0, bytesRead);
                }
                byte[] result = stream.ToArray();
                var m = new Memory<byte>(result);
                await cnt.Response.BodyWriter.WriteAsync(m);
            });
        });
    }

}

4. That is all

Windows terminal + Powershell to run IDE, Angular, .NET Core

I work at http://github.com/ignatandrei/BlocklyAutomation – and every time I need to run Angular , Visual Studio Code IDE and .NET Core run ( and Visual Studio sometimes – it is more web based)

Also, sometimes I need to run in a container – to install globally something that I do not want to have in my PC.

So I have 1 windows terminal command to run those :

First one

wt new-tab –title RootSource –suppressApplicationTitle -p “Windows PowerShell” -d . cmd /k “cd src && code . && powershell” ;split-pane –title Angular –suppressApplicationTitle -V -p “Windows PowerShell” -d . cmd /k “cd src && npm run start –watch && powershell” ;split-pane –title LocalAPI_NetCore –suppressApplicationTitle -V -p “Windows PowerShell” -d . cmd /k “cd src/Local/LocalAPI/LocalAPI && dotnet watch run”

As you see , I have

code .

to run VSCode,

npm run start –watch

to run Angular and

dotnet watch run

to run dotnet.

Second one:

wt new-tab -p “Windows PowerShell” -d . ;split-pane -p “Windows PowerShell” -d . cmd /k “cd src && devcontainer open .”

The second it just works with devcontainer – if you are interested in sources, see https://github.com/ignatandrei/BlocklyAutomation/tree/main/src/.devcontainer

With those, I can start my developer environment fast ( how fast, that depends on my PC)

Python vs C#

I have read the article https://www.red-gate.com/simple-talk/development/dotnet-development/10-reasons-python-better-than-c-sharp

I think that deserves an answer – so I take all points to answer to them. You can replicate the code in C# 10 / .NET 6 with a globalusings.cs with the following content:


global using static System.Console;
global using static System.Linq.Enumerable;
global using static System.IO.Directory;
global using static System.String;
global using static System.Math;

So let’s start:

Point 1 - Indented code blocks


//Point 1 - Indented code blocks
//With curly braces I can indent whatever way I want - I am not constrained

using System.Collections.Generic;

foreach (var i in Range(1, 11))
{
if (i % 3 == 0)
WriteLine("fizz");
else if (i % 5 == 0)
WriteLine("buzz");

}

Also , when I want to comment something , with C# I know when the else if is
terminating without counting tabs. Just see where the { is closing with }

Point 2 -  Self-declaring Variables

Seriously? What is the difference between Python
//meaning_of_life = 42
and C# ? Just a var ?

//Point 2 - Self-declaring Variables
var meaning_of_life = 42;

Point 3 -– Those modules

It is the same code as Python

foreach (var file in EnumerateFiles(@"C:\")){
 WriteLine(file);
}

Point 4 – Simplicity of Data Structures

Quoting : Purists will say that you need all the different data types that C# provides, but trust me, you don’t.
I want to think about how to modelate a CallCenter waiting for phones – Stack or Queue ?
I want to think about how to modelate a CallCenter waiting for phones with business consumer that has priority ?https://docs.microsoft.com/en-us/dotnet/api/system.collections.generic.priorityqueue

Point 5 - Slicing

Same code in C#

var sins = new[]{ "pride", "envy", "gluttony", "greed", "lust", "sloth", "wrath" };
//var selected_sins = sins[1..(sins.Length-1)];
var selected_sins = sins[1..^1];
WriteLine(Join(" ", selected_sins));

Point 6 – For loops

Same syntax in C#

foreach (var i in Range(0,5))
{
WriteLine(i);
}
// words in a string
string[] words = { "Simple", "Talk" };
foreach (string word in words)
{
WriteLine(word);
}

Quoting : “(unlike in C#, where sometimes you use for and sometimes foreach)”
Please enumerate an array backwards in python without reversing

7 – List comprehensions

This is similar code in C#

var nr = Range(1, 11).Where(it => it % 2 == 0).Select(it => Pow(it, 3));
WriteLine(Join(" ", nr));

8 – Sets

Same example

//this contains some duplicates
var languages = new[] { "C#", "Python", "VB", "Java", "C#", "Java", "C#" };
//more dense than Python
languages = new HashSet<string>(languages).ToArray();
//this will give: ['C#', 'Java', 'Python', 'VB']
WriteLine(Join(" ", languages));

Second example

//create two sets: Friends characters and large Antarctic ice shelves
var friends = new[] { "Rachel", "Phoebe", "Chandler", "Joey", "Monica", "Ross" };
var ice_shelves = new[] { "Ronnie-Filchner", "Ross", "McMurdo" };
// show the intersection (elements in both lists)
WriteLine(Join(" ", friends.Intersect(ice_shelves)));
//show the union (elements in either list)
WriteLine(Join(" ", friends.Union(ice_shelves)));
// show the friends who aren't ice shelves
WriteLine(Join(" ", friends.Except(ice_shelves)));
// elements in either set but not both -- more difficult
WriteLine(Join(" ", friends.Union(ice_shelves).Except(friends.Intersect(ice_shelves))));

9 – Working with files and folders

//# list of 3 people
var people = new[] { "Shadrach", "Meshach", "Abednego" };
//# write them to a file
File.WriteAllLines(@"C:\all\a.txt", people);

10 – The quality of online help

See https://docs.microsoft.com/ – good tutorials

FileExtension–part 6 – CI/CD

So now how to make visible all those coding ?

For tests –CodeCov, https://app.codecov.io/gh/ignatandrei/FileExtension hosts for free to see the coverage

For documentation – ReadTheDocs, https://fileextension.readthedocs.io/en/latest/ hosts for free.

For packaging – NuGEt https://www.nuget.org/packages/FileExtension/

For more detailed analysis Sonar https://sonarcloud.io/summary/overall?id=ignatandrei_FileExtension can show code smells and bugs and more

GitHub actions https://github.com/ignatandrei/FileExtension/actions allows me to run all those automatically and gather results

For showing a website – Azure – https://fileextension.azurewebsites.net/ can show the usage.

I was thinking that it is enough for the people to understand the application

FileExtension–Tests–part 5

The tests are the easy part. I have put some files on a folder( https://github.com/ignatandrei/FileExtension/tree/master/src/TestFileExtensions/TestFiles ) and test if the extension matches the recognizer.

As NuGet packages I used

FluentAssertions

xunit

LightBDD.XUnit2

FluentAssertion allows me to write more readable tests , like

AllExtensions.Should().HaveCountGreaterThanOrEqualTo(number);

LightBDD allows me to write tests in a more “business “ way :

await Runner
     .AddSteps(Given_The_Recognizer)
     .AddAsyncSteps(_ => When_Read_The_File(fileName))
     .AddSteps(
         _ => Then_Should_Recognize_File(fileName),
         _ => Then_The_Extension_Matches(fileName)
     )
     .RunAsync();

You can see the result at https://fileextension.readthedocs.io/en/latest/BDD/LightBDDReport/ = go down to see the Given / When /

Also, for enumerating files from the folder to test, I use ClassData

[ClassData(typeof(DirectoryTestData))]
public async void TestMultipleFiles(string nameFile)

Category: .NET Core

Dependent Framework Versioning

For .NET Core 5– https://www.nuget.org/packages/NetCore2Blockly/5.2022.210.2007

Andrei Ignat weekly software news(mostly .NET)