DEV Community: Mateusz Jarzyna

fkill - (imo) best way to kill processes // Cool GitHub projects

Mateusz Jarzyna — Sat, 19 Sep 2020 18:53:57 +0000

I will start with the biggest disadvantage of this app - fkill is written in JS using Node.js. To be clear - JS is not my enemy, I'm not using that stack, so I had to install npm and so on.

sindresorhus / fkill-cli

Fabulously kill processes. Cross-platform.

Fabulously kill processes. Cross-platform.

Works on macOS, Linux, and Windows.

Install

$ npm install --global fkill-cli

Usage

$ fkill --help

  Usage
    $ fkill [<pid|name> …]

  Options
    --force -f    Force kill
    --verbose -v  Show process arguments
    --silent -s   Silently kill and always exit with code 0

  Examples
    $ fkill 1337
    $ fkill safari
    $ fkill :8080
    $ fkill 1337 safari :8080
    $ fkill

  To kill a port, prefix it with a colon. For example: :8080.

  Run without arguments to use the interactive interface.
  In interactive mode, 🚦n% indicates high CPU usage and 🐏n% indicates high memory usage.

  The process name is case insensitive.

Interactive UI

Run fkill without arguments to launch the interactive UI.

fkill - API for this module
alfred-fkill - Alfred workflow for this module

Maintainers

View on GitHub

Install

npm install --global fkill-cli

Features

kill process by pid - fkill 1337 (nothing fancy)
kill process by name - fkill firefox (like killall)
kill process by port - fkill :8080 - will kill process which uses port 8080, super cool feature
interactive mode - fkill - shows all processes, search bar, ports used by each process, memory usage and CPU usage
-f to force kill

LazyGit - terminal UI for git commands // Cool GitHub projects

Mateusz Jarzyna — Sat, 19 Sep 2020 12:19:57 +0000

Are you sad when you are typing git add shitty/long/path/to/file.exe, but you can't use SourceTree (or other GUI) because you are proud terminal user?

I was too. But now I'm using LazyGit!

jesseduffield / lazygit

simple terminal UI for git commands

^{Special thanks to:}

Warp is a modern, Rust-based terminal with AI built in so you and your team can build great software, faster.

^{Visit warp.dev to learn more.}

I (Jesse) co-founded Subble to save your company time and money by helping you manage its software subscriptions. Check it out!

A simple terminal UI for git commands

Elevator Pitch

Rant time: You've heard it before, git is powerful, but what good is that power when everything is so damn hard to do? Interactive rebasing requires you to edit a goddamn TODO file in your editor? Are you kidding me? To stage part of a file you need to use a command line program to step…

View on GitHub

Installation

You can install it on MacOS using brew

brew install jesseduffield/lazygit/lazygit

(Linux and Widows are also supported, just read the manual)

Configuration

LazyGit, as every good terminal app, is configurable. Documentation is not the best, but good enough. For example, you can use it with white theme

About series

Cool GitHub projects is a blog posts series where I will share useful/fancy projects I found on GitHub

10 Intellij Idea's plugins I'm using

Mateusz Jarzyna — Tue, 09 Jun 2020 16:03:35 +0000

^{Cover image: https://pl.m.wikipedia.org/wiki/Plik:IntelliJIDEA_2016.3_Community.png}

Intellij Idea has a lot of plugins, both paid and free (I'm using only free). Here is a list of my favorite, that I'm using daily.

^{All images are from the official plugin's web page}

SonarLint

SonarLint is a static code analyzer. Idea has build-in code analyzer, but SonarLint is able to find more issues in your Java code, so they work perfectly together. SonarLint can be connected to SonarQube server, but even in standalone is an interesting plugin.

SonarLint plugin page

Rainbow Brackets

Rainbow Brackets is just an eye candy tool that makes brackets colorful. There is also Indent Rainbow but I'm not using it.

Rainbow Brackets plugin page

Code review tool

That's what I'm most happy about. Making code review in web browser is a pain, you can't see the full code, you can't go to method implementation in simply way, etc. Moving this process to the IDE was fantastic decision.

In my company we're using the GitLab, so I'm using Merge Request Integration CE - Code Review for GitLab. But if you are using GitHub or BitBucket or other tool for code review - I'm sure that there is a good plugin for you, just search and test.

Extra Icons

Extra Icons adds icons for popular tools that are not available in Idea by default.

Extra Icons plugin page

CSV Plugin

CSV Plugin allows to open the .csv files in tabular, excel-like view. Useful from time to time.

CSV Plugin plugin page

CodeGlance

CodeGlance adds a code mini map into the editor pane. Not very useful, just show yours code structure.

CodeGlance plugin page

CamelCase

[no picture]
CodeGlance allows switching easily between kebab-case, SNAKE_CASE, PascalCase, camelCase, snake_case or space case. I'm using it when renaming the variable after changing the scope.

CodeGlance plugin page

AsciiDoc

In my company we are using Spring REST Docs to create technical documentation. AsciiDoc plugin just shows rendered preview of .adoc files.

AsciiDoc plugin page

Request mapper

Request mapper is a tool that allows you to find a URL mapping declaration in your Spring/JAX-RS/Micronaut project. You can add @RequestMapping annotation on both, class and method and the plugin still works fine and quick.

Request mapper plugin page

IDE Features Trainer

I've used this plugin in the past. As the home page says:

Learn basic shortcuts and essential features interactively — right inside the IDE.

You can learn the IDE in an easy, interactive way.

IDE Features Trainer plugin page

Bonus 1 - the font

I'm using the Monaco font. In my opinion it's really readable and clean. I love that font.

Bonus 2 - Live templates

I'm using only one live template in my daily job - to create new JUnit test.
To create that one, open Preferences ⇾ Editor ⇾ Live Templates, then add new template and paste:

@org.junit.jupiter.api.Test
public void $END$() throws Exception {

}

Bonus 3 - skip Spring's classes when debugging

I was really annoyed when I was debugging my application and saw infinite numbers of Spring's classes. So I've opened Preferences ⇾ Build, Execution, Deployment ⇾ Debugger ⇾ Stepping and added org.springframework.* into the Do not step into the clasess list. Really helpful trick ;)

Share in comments yours favorite plugins and tricks.

MAKERbuino - solder your own game console and start making games

Mateusz Jarzyna — Mon, 30 Mar 2020 17:57:07 +0000

I've been the software developer for several years now. But in this profession there was sill an area in which I have exactly 0 experience - gamedev. I know I can watch tutorials about the Unity and build 3d games in two weeks. But to tell you the truth - it's not what I like. I like to know how it exactly works, not only how to use the framework.
Modern games engines are really, really too complicated to understand them. Even games from '90 are too complicated for me - especially when I have no experience in game development, and also too little in C language.
So I decided to start with the basic. How the games works on yours old, two-color phone? Simple gameplay, low graphic, poor hardware - I think it's the good point to start the journey.

Soldering the MAKERbuino

There is a toy on the market called MAKERbuino. Actuallt it's Gamebiono Classic that you can solder by yourself.

The box contains few resistors, some buttons, one old screen and so on.

I'm not sure why my console was delivered with connected battery, but well, doesn't matter, it works anyway.
To be honest - I'm really hopeless in soldering. But hey, practice makes perfect!
So I prepared all the necessary tool...

... and I started soldering

I'm such a noob I didn't realize I was soldering with the battery connected!
The official guide is really user-friendly. It says that it should take 5 hours, I've done it in 4.

Not perfect, not terrible. The most important thing is that it works.
Four hours and one liter of water later finally I've soldered the console.

Sound works, screen brightness works, old Nokia's screen works (and can display exactly one color, it's weird nowadays) works. The MAKERbuino comes with SD card (128 MB :), inside there were few build-in games.

WOW, everything works, I'm so proud of myself.
There is only one big shortcoming - the "click" sound, it is sooo loud...

I'm game developer!

ATmega328p-pu - 2kB of RAM and an 8-bit CPU at 16MHz. I've programmed few times on so poor computer (for example I made a 3d printed fish feeder with Adruino), but sill - nowadays it's a little challenge, I can't use a dozen of frameworks. Also, I have to use skills sightly forgotten by modern programmers - optimization.
MAKERbuino is built on Arduino, so I had to configure my Arduino IDE. Fortunately official guide explains very well how to do it. Arduino IDE is not as good as Intellij Idea for example, it's very poor to be honest. But to create 100 lines of code it's good enough.

So, how to create a game? I don't know. There is no tutorial, no API. But MAKERbuino is after all Gamebuino Classic. So I used the Gamebuino tutorials. Official Gabebuino Academy is nice place to start. But it uses the new Gamebuino - with colorful screen and little different API - so I had to forgot about coping-and-pasting. I read those tutorials to understand the process and start thinking like a game developer. I've also had to use old reference to write the code.

Bouncing ball

Uploading the game to the console is not the fastest process in the world. And that loud sound of clicking. Fortunately there is a online emulator.
One hour after reading the Academy I've made bouncing ball

I have no idea what is that weird sign in right-top corner, doesn't matter. The ball is bouncing!

Full source code:

#include <Gamebuino.h>

Gamebuino gb;

int ballX = LCDWIDTH  /2;
int ballSpeedX = 1;
int ballY = LCDHEIGHT /2;
int ballSpeedY = 1;

void setup() {
  gb.begin();

  gb.titleScreen("bouncing ball");
}    

void loop() {
  while(!gb.update());
  gb.display.clear();

  if (ballX == 0) {
    ballSpeedX = 1;
  } else if (ballX == LCDWIDTH - 2) {
    ballSpeedX = -1;
  }

  if (ballY == 0) {
    ballSpeedY = 1;
  } else if (ballY == LCDHEIGHT - 2) {
    ballSpeedY = -1;
  }

  ballX += ballSpeedX;
  ballY += ballSpeedY;
  gb.display.fillRect(ballX, ballY, 2, 2);
}

Maybe I can optimize and refactor the code, but hey, it's my first "game" ever.

PONG

One or two hours later I've made probably the simplest game ever - the PONG. With "Artificial Intelligence"! The CPU player is very simple of course

If ball is higher than paddle
   move up
If ball is lower than paddle
   move down

The ball was much bigger on the emulator.

Full source code:

#include <Gamebuino.h>

Gamebuino gb;

int ballX = 0;
int ballSpeedX = 0;
int ballY = 0;
int ballSpeedY = 0;
int ballSize = 2;

int paddleWidth = 3;
int paddleHeight = 15;

int playerPaddleY = 15;
int playerPaddlePadding = 3;
int cpuPaddleY = 15;
int cpuPaddlePadding = LCDWIDTH - paddleWidth - 3;

int playerScore = 0;
int cpuScore = 0;

void resetBall() {
  ballX = LCDWIDTH  /2;
  ballY = LCDHEIGHT /2;
  if (random(0, 2) == 0) {
    ballSpeedX = 1;
  } else {
    ballSpeedX = -1;
  }
  if (random(0, 2) == 0) {
    ballSpeedY = 1; 
  } else {
    ballSpeedY = -1;
  }
}

void setup() {
  gb.begin();

  gb.titleScreen("super gra");
  resetBall();
}    

void loop() {
  while(!gb.update());
  gb.display.clear();

  if (ballX == 0) {
    // player lose
    cpuScore++;
    resetBall();
  } else if (ballX == LCDWIDTH - ballSize) {
    // cpu lose
    playerScore++;
    resetBall();
  }

  if (ballY == 0) {
    ballSpeedY = 1;
  } else if (ballY == LCDHEIGHT - ballSize) {
    ballSpeedY = -1;
  }

  if (ballX == playerPaddlePadding + paddleWidth
      && ballY >= playerPaddleY
      && ballY <= playerPaddleY + paddleHeight) {
    // touch player's paddle
    ballSpeedX = 1;
  } else if (ballX == cpuPaddlePadding - ballSize
      && ballY >= cpuPaddleY
      && ballY <= cpuPaddleY + paddleHeight) {
    // touch cpu's paddle
    ballSpeedX = -1;
  }

  ballX += ballSpeedX;
  ballY += ballSpeedY;
  gb.display.fillRect(ballX, ballY, ballSize, ballSize);


  if (gb.buttons.repeat(BTN_UP, 0) && playerPaddleY > 0) {
    playerPaddleY -= 1;
  }
  if (gb.buttons.repeat(BTN_DOWN, 0) && playerPaddleY < LCDHEIGHT - paddleHeight) {
    playerPaddleY += 1;
  }

  int cpuPaddleCenter = (cpuPaddleY + paddleHeight) / 2;
  if (cpuPaddleCenter > ballY && cpuPaddleY > 0) {
    cpuPaddleY--;
  } else if (cpuPaddleCenter < ballY && cpuPaddleY < LCDHEIGHT - paddleHeight) {
    cpuPaddleY++;
  }

  gb.display.fillRect(playerPaddlePadding, playerPaddleY, paddleWidth, paddleHeight);
  gb.display.fillRect(cpuPaddlePadding, cpuPaddleY, paddleWidth, paddleHeight);

  gb.display.cursorY = 3;
  gb.display.cursorX = 10;
  gb.display.print(playerScore);
  gb.display.cursorX = LCDWIDTH - 13;
  gb.display.print(cpuScore);
}

Now I see that I should refactor my code. But I'm satisfied and so proud of myself.

A lot of fun

New experiences are always full of a fun, in this case a lot of fun. Soldering, making game from '70 - I had a nice time. Maybe it's not suitable for CV, but it's still a lot of experience.
If you are a fan of electronic - I can recommend from the bottom of my heart. It's time to finish this post and finish "snake" game. And later maybe Micro Machines, who knows...

Build your own HTTP server in Java in less than one hour (only GET method)

Mateusz Jarzyna — Wed, 25 Mar 2020 22:07:29 +0000

One of the most frequency used protocol in the whole Internet *

* In OSI model, layer 7

Every time you visit a website your web browser uses the HTTP protocol to communicate with web server and fetch the page's content. Also, when you are implementing backend app and you have to communicate with other backend app - 80% (or more) of cases you will use the HTTP.

Long story short - when you want to be a good software developer, you have to know how the HTTP protocol works. And wiring the HTTP server is pretty good way to understood, I think.

What a web browser sends to the web server?

Good question. Of course, you can use "developer tools", let's do it.

Hmm... but what now? What exactly it means? We can see some URL, some method, some status, version (huh?), headers, and other stuff. Useful? Yes, but only to analyze the web app, when something is wrong. We still don't know how HTTP works.

Wireshark, my old friend

The source of truth. Wireshark is application to analyze network traffic. You can use it to see each packet that is sent by your (or to your) PC.

But to be honest - if you know how to use Wireshark - you probably know how HTTP and TCP works. It's pretty advanced program.

You are right - the specification

Every good (I mean - used by more that 5 peoples) protocols should have specification. In this case it's called RFC. But don't lie - you will never read this, it's too long - https://tools.ietf.org/html/rfc2616 .

Just run the server and test

Joke? No. Probably you have installed on your PC very powerful tool called netcat, it's pretty advanced tool.
One of the netcat features is TCP server. You can run the netcat to listen on specific port and print every thing what it gets. Netcat is a command line app.



nc -v -l -p 8080

Netcat (nc), please, listen (-l) on port 8080 (-p 8080) and print everything (-v).

Now open web browser and enter http://localhost:8080/. Your browser will send the HTTP request to the server runned by netcat. Of course nc will print the whole request and ignore it, browser will wait for the response (will timeout soon). To kill nc press ctrl+c.

So, finally, we have an HTTP request!



GET / HTTP/1.1
Host: localhost:8080
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:74.0) Gecko/20100101 Firefox/74.0
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8
Accept-Language: en-US,en;q=0.5
Accept-Encoding: gzip, deflate
Connection: keep-alive
Cookie: JSESSIONID=D3AF43EBFC0C9D92AD9C37823C4BB299
Upgrade-Insecure-Requests: 1

As you can see - it fully texts protocol. No bits to analyze, just plain text.

HTTP request

It may be a little confusing. Maybe nc parses the request before printing? HTTP protocol should be complicated, where is the sequence of 0 and 1? There aren't any. HTTP is really very simple text protocol. There is only one, little trap (I will explain it at the end of this section).

We can split the request to the 4 main parts:



GET / HTTP/1.1

This is the main request.

GET - this is the HTTP method. Probably you know there are a lot of methods.
GET means give me

/ - resource. / means default one.
When you will open localhost:8080/my_gf_nudes.html, the resource will be /my_gf_nudes.html

HTTP/1.1 - HTTP version. There are few versions, 1.1 is commonly used.



Host: localhost:8080

Host. One server can host many domains, using this field, the browser says which domain it wants exactly



User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:74.0) Gecko/20100101 Firefox/74.0
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8
Accept-Language: en-US,en;q=0.5
Accept-Encoding: gzip, deflate
Connection: keep-alive
Cookie: JSESSIONID=D3AF43EBFC0C9D92AD9C37823C4BB299
Upgrade-Insecure-Requests: 1

Headers. In short: some additional informations. But I'm sure that you know what headers are :)

Surprise - empty line. It means: end of the request. In general - empty line in HTTP means end of section.

The trap

In HTTP every new line separator is a Window's new line. \r\n not \n. Remember.

Response

Ok. We have a request. How does response look like? Send a request to any server and see, there is nothing simpler.
On your laptop you can find another very useful tool - telnet. Using telenet you can open TCP connection, write something to server and print the response.
Try to do it yourself. Run telnet google.com 80 (80 is the default HTTP port) and type request manually (you know how it should look like). To close connection press ctrl+] than type quit.

OK. We have a response.



HTTP/1.1 301 Moved Permanently
Location: http://www.google.com/
Content-Type: text/html; charset=UTF-8
Date: Wed, 25 Mar 2020 18:53:12 GMT
Expires: Fri, 24 Apr 2020 18:53:12 GMT
Cache-Control: public, max-age=2592000
Server: gws
Content-Length: 219
X-XSS-Protection: 0
X-Frame-Options: SAMEORIGIN

<HTML><HEAD><meta http-equiv="content-type" content="text/html;charset=utf-8">
<TITLE>301 Moved</TITLE></HEAD><BODY>
<H1>301 Moved</H1>
The document has moved
<A HREF="http://www.google.com/">here</A>.
</BODY></HTML>

We can split it to 4 sections



HTTP/1.1 301 Moved Permanently

HTTP/1.1 - version
301 - status code. I believe you are familiar with that
Moved Permanently - human-readable status code



Location: http://www.google.com/
Content-Type: text/html; charset=UTF-8
Date: Wed, 25 Mar 2020 18:53:12 GMT
Expires: Fri, 24 Apr 2020 18:53:12 GMT
Cache-Control: public, max-age=2592000
Server: gws
Content-Length: 219
X-XSS-Protection: 0
X-Frame-Options: SAMEORIGIN

Headers

Empty line, it means that the content will be sent in next section.



<HTML><HEAD><meta http-equiv="content-type" content="text/html;charset=utf-8">
<TITLE>301 Moved</TITLE></HEAD><BODY>
<H1>301 Moved</H1>
The document has moved
<A HREF="http://www.google.com/">here</A>.
</BODY></HTML>

Content, HTML or binary or something

Empty line, means end of request.

REMEMBER: each new line is \r\n

Time for programming!

We know how request look like, we know how response look like, it's time to implement our server.

What we expect

We want to get a very simple thing - to display an HTML page and a picture in a browser.
Let's prepare two HTMLs files and one picture



❯ pwd
/tmp/www

❯ ls
gallery.html index.html   me.jpg

❯ cat index.html
<html>
  <header>
    <title>My homepage!</title>
  </header>
  <body>
    <h1>Welcome!</h1>
    <p><a href="gallery.html">Here</a> you can look at my pictures</p>
  </body>
</html>

❯ cat gallery.html
<html>
  <head>
    <title>Gallery</title>
  </head>
  <body>
    <h1>My sexi photos<h1>
    <img src="me.jpg" />
  </body>
</html>

❯

The plan

Plan is very simple:

Open TCP socket and listen
Accept the client and read request
Parse the request
Find requested resource on disk
Send the response
Test

Open TCP socket

In this article we will use ServerSocket class to handle TCP connection. As a homework you can reimplement the server to use the classes from the nio packages.

So, open your IDE and let's start.



    public static void main( String[] args ) throws Exception {
        try (ServerSocket serverSocket = new ServerSocket(8080)) {
            while (true) {
                // implement client handler here
            }
        }
    }

I want to keep the code concise and clean - it's why I throws Exception instead of implementing good exception handling.
So as I told, we have to open socket on port 8080 (why not 80? Because to use low port you need root privileges).
We also need the infinity loop to 'pause the server'.

User telnet to test the socket:

Perfect, works.

Accept client connection



        try (ServerSocket serverSocket = new ServerSocket(8080)) {
            while (true) {
                try (Socket client = serverSocket.accept()) {
                    handleClient(client);
                }
            }
        }

To accept connection from client we have to call blocking accept() method. Java program will wait for a client on that line.

Time to implement the client handler:



    private static void handleClient(Socket client) throws IOException {
        System.out.println("Debug: got new client " + client.toString());
        BufferedReader br = new BufferedReader(new InputStreamReader(client.getInputStream()));

        StringBuilder requestBuilder = new StringBuilder();
        String line;
        while (!(line = br.readLine()).isBlank()) {
            requestBuilder.append(line + "\r\n");
        }

        String request = requestBuilder.toString();
        System.out.println(request);
    }

We have to read the request. How? Just read the input stream from the client's socket. In Java it's not so simple, that's why I made this ugly line



new BufferedReader(new InputStreamReader(client.getInputStream()));

Well, Java.

Request ends with one empty line (\r\n), remember? Client will send empty line, but imputStream will be still open, we have to read it until one, empty line arrives.

Run the server, go to http://localhost:8080/ and observe logs:

It works! We can log the whole request!

Parse the request

Parsing the request is realy simple, I don't think there's any need for further explanation



        String request = requestBuilder.toString();
        String[] requestsLines = request.split("\r\n");
        String[] requestLine = requestsLines[0].split(" ");
        String method = requestLine[0];
        String path = requestLine[1];
        String version = requestLine[2];
        String host = requestsLines[1].split(" ")[1];

        List<String> headers = new ArrayList<>();
        for (int h = 2; h < requestsLines.length; h++) {
            String header = requestsLines[h];
            headers.add(header);
        }

        String accessLog = String.format("Client %s, method %s, path %s, version %s, host %s, headers %s",
                client.toString(), method, path, version, host, headers.toString());
        System.out.println(accessLog);

Just some splits. The only one thing that you may not understand is why we started the loop from 2? Because first line (index 0) is GET / HTTP/1.1, second line is host. The headers start with the third line of the request

Send response

We will send the response to the client's output stream.



        OutputStream clientOutput = client.getOutputStream();
        clientOutput.write("HTTP/1.1 200 OK\r\n".getBytes());
        clientOutput.write(("ContentType: text/html\r\n").getBytes());
        clientOutput.write("\r\n".getBytes());
        clientOutput.write("<b>It works!</b>".getBytes());
        clientOutput.write("\r\n\r\n".getBytes());
        clientOutput.flush();
        client.close();

Do you remember how response should look like?



version status code
headers
(empty line)
content
(empty line)

Don't forget to close the output stream.

Wow, it's really works

Find requested resource

We have to implement two methods first



    private static String guessContentType(Path filePath) throws IOException {
        return Files.probeContentType(filePath);
    }

    private static Path getFilePath(String path) {
        if ("/".equals(path)) {
            path = "/index.html";
        }

        return Paths.get("/tmp/www", path);
    }

guessContentType - we have to tell to the browser what kind of content we are sending. It's callend content type. Fortunately, there are built-in mechanisms in Java for this. We don't have to make a big switch block.

getFilePath - Before we will return the file - we need to known it location.
This condition deserves attention



        if ("/".equals(path)) {
            path = "/index.html";
        }

If user wants default resource then return index.html.

send the response

Do you remember the code that sends response to the user (block of clientOutput.write)? We need to move it to the method



    private static void sendResponse(Socket client, String status, String contentType, byte[] content) throws IOException {
        OutputStream clientOutput = client.getOutputStream();
        clientOutput.write(("HTTP/1.1 \r\n" + status).getBytes());
        clientOutput.write(("ContentType: " + contentType + "\r\n").getBytes());
        clientOutput.write("\r\n".getBytes());
        clientOutput.write(content);
        clientOutput.write("\r\n\r\n".getBytes());
        clientOutput.flush();
        client.close();
    }

Ok, finally we can return the file



        Path filePath = getFilePath(path);
        if (Files.exists(filePath)) {
            // file exist
            String contentType = guessContentType(filePath);
            sendResponse(client, "200 OK", contentType, Files.readAllBytes(filePath));
        } else {
            // 404
            byte[] notFoundContent = "<h1>Not found :(</h1>".getBytes();
            sendResponse(client, "404 Not Found", "text/html", notFoundContent);
        }

It works!

Finally, we can see html page served by our web server!

Homework

Make it multi-thread.
1. Create thread pool
2. Move handleClient method to separated class and run it in new thread
Rewrite it using non-blocking IO
Implement POST method
1. Start netcat
2. Send some HTML form
3. Analyze request

Full source code

Linux's commands and tricks I'm using in my daily job as a developer

Mateusz Jarzyna — Mon, 11 Nov 2019 10:50:35 +0000

This is not a post from the series of those describing the cd command. It's just a list of commands and tricks I'm using (almost) every day.

Port forwarding

Sometimes I have to connect to database and of course I prefer to use my GUI manager (JetBrains DataGrip).
So, if security policy exist in your company and your database's port is not exposed you can execute

ssh -L{port on your PC}:localhost:{database's port} root@{server IP}

The command below will open port 3308 on your laptop and everything will be forwarded to 192.168.1.2:3306

ssh -L3308:localhost:3306 root@192.168.1.2

localhost means that database is listening on 192.168.1.2. You can type for example 192.168.3.77 and everything will be forwarded to .3.77 server via .1.2.

Edit file in VIM without `sudo`, but save with `sudo`

Have you ever edited some configs file and forgot to sudo? Me too... There is a trick to save the file anyway, just type in VIM:

:w !sudo tee %

Explanation

Go to beggining/end of line in terminal

If you wrote a very long command in the terminal it may take a long time before you return to the begging of the line to add missing sudo. And back to the end to add some parameters.
Press crtl + a to move to the begging and crtl + e to the end of the line in terminal.

ll

Save few days in a year by typing ll instead of ls -la. Works on most Linux servers.

Execute command you executed in the past

Last command

To execute last command over again you can of course press ↑ (arrow up) key. But you can also type !!. So executing last command as a root is very easy

sudo !!

To run the last command that started with apt type !apt

Search history

To find the command that contains /tmp you have executed in the past press ctrl + r and type /tmp. Press ctrl + r again for next result.
To show all commands or to search using regular expression use

history | grep "/tmp"

Agree for everything

To say yes for each question you can use application called yes

yes | yum install curl

use yes no to say no and discard.

WARNING

As @patricnox notices in the comment - using yes may do unexpected things. You can accidentally install 10 GB of dependencies or other things you don't want to do.

Run a long-lasting process in the background and close the terminal

If you run a script that will end in 3 days, you don't have to wait with the terminal window open to end. You can run it using nohup command

nohup wget http://large-files.com/10gb-super-movie.avi &

wget works in the background, output is saved to nohup.out file in working directory.

Checking who has stolen your favourite port

It's really annoying when you are trying to run nginx but you can't because there is already apache running and port 443 is busy.
So, how to determinate which process is listening on port 80:

$ netstat -tulpn | grep 80
tcp6       0      0 :::80                 :::*                   LISTEN     10177/java

10177 is a pid you are looking for. Now execute

ps aux | grep 10177

for more details.

Reading logs

Everyone knows that less is a very good way to read a logs files. But you can also read gziped logs without extracting!

less /var/log/my-app/my-app.log.2015.12.14.gz

Live reading

tail -f /var/log/my-app/my-app.log | grep ERROR

The command above will show only new lines that contains ERROR.

Sort process

Show top 3 processes sorted by CPU usage

ps aux --sort=-pcpu | head -n 4

Show top 3 processes sorted by memory usage

ps aux --sort=-rss | head -n 4

Executing command every X seconds

To print command's output every X seconds you can use watch command. For example to create clock run

watch -n 1 date

Quiet mode

A lot of standards commands has quiet or silent mode. Very useful when you are creating some bash script. In most of the cases just add -q or -s (read --help or man or check on StackOverflow)

zip -q archive.zip big-file.jpg

But sometimes (practically always with in-house scripts) you have to ignore the output (send to /dev/null)

./very-verbose.sh 1>/dev/null

Create log files for scripts executed by crontab

0 22 * * 1-5 /opt/scripts/send-report.sh 2>/var/log/scripts/report-error.log

So next time when your script will fail you won't lose the reason

DEV Community: Mateusz Jarzyna

fkill - (imo) best way to kill processes // Cool GitHub projects

sindresorhus / fkill-cli

Fabulously kill processes. Cross-platform.

Install

Usage

Interactive UI

Related

Maintainers

Install

Features

LazyGit - terminal UI for git commands // Cool GitHub projects

jesseduffield / lazygit

simple terminal UI for git commands

Sponsors

Elevator Pitch

Installation

Configuration

About series

10 Intellij Idea's plugins I'm using

SonarLint

Rainbow Brackets

Code review tool

Extra Icons

CSV Plugin

CodeGlance

CamelCase

AsciiDoc

Request mapper

IDE Features Trainer

Bonus 1 - the font

Bonus 2 - Live templates

Bonus 3 - skip Spring's classes when debugging

MAKERbuino - solder your own game console and start making games

Soldering the MAKERbuino

I'm game developer!

Bouncing ball

PONG

A lot of fun

Build your own HTTP server in Java in less than one hour (only GET method)

One of the most frequency used protocol in the whole Internet *

What a web browser sends to the web server?

Wireshark, my old friend

You are right - the specification

Just run the server and test

HTTP request

The trap

Response

Time for programming!

What we expect

The plan

Open TCP socket

Accept client connection

Parse the request

Send response

Find requested resource

send the response

It works!

Homework

Full source code

Linux's commands and tricks I'm using in my daily job as a developer

Port forwarding

Edit file in VIM without sudo, but save with sudo

Go to beggining/end of line in terminal

ll

Execute command you executed in the past

Last command

Search history

Agree for everything

Run a long-lasting process in the background and close the terminal

Checking who has stolen your favourite port

Reading logs

Live reading

Sort process

Executing command every X seconds

Quiet mode

Create log files for scripts executed by crontab

Edit file in VIM without `sudo`, but save with `sudo`