This article was originally published here

Intro

I’ve always loved everything to do with Telegram, and since I’m studying programming, I couldn’t pass up the opportunity to make my own bot. And it is quite easy to do simple things, there is a lot of documentation.

Requirements

We do not need much to achieve our purpose, I already announced at the beginning that it is really easy. We will need:

• A telegram account: necessary to create our bot.
• Install Node.js: is what we will use to program our bot.
• Create an account in railway.app: there we will host our bot. It is highly recommended that you use a Github repository for the code, that way with each commit your bot will be updated automatically.

First steps

About the first steps there is quite a lot of information on the internet, as they are the same regardless of the programming language you decide to use: you must create your bot with @BotFather. After typing /newbot, you make some decisions such as the profile picture, the description…

Here the most important thing is to save the token returned by the bot, since the plugin we will use (Telebot) needs it.

Writing our code

This is the fun part. We have to create a folder for our project and in it open some terminal (in Windows you can do it by pressing shift + right click and selecting “Open powershell window here”).

With npm init and completing the questions, we are ready to start writing the code for our bot. We will not close the terminal yet.

Now we will create 2 new files: bot.js y .gitignore, the first one for the bot code itself and the second one to avoid uploading folders like node_modules to github.

Precisely in .gitignore we will write

node_modules
package-lock.json

Now we have to install Telebot. Type npm install telebot in the terminal and wait for it to finish.

In bot.js we type:

 import TeleBot from "telebot";

const bot = new TeleBot({
  token: process.env.TG_TOKEN,
  polling: {
    // Optional. Use polling.
    interval: 1000, // Optional. How often check updates (in ms).
    timeout: 0, // Optional. Update polling timeout (0 - short polling).
    limit: 3, // Optional. Limits the number of updates to be retrieved.
    retryTimeout: 5000, // Optional. Reconnecting timeout (in ms).
  },
  usePlugins: ["commandButton", "reporter"],
  pluginConfig: {
    reporter: {
      // What to report?
      events: ["reconnect", "reconnected", "stop", "error"],
      // User list
      to: [my_id],
    },
  },
});

bot.on(/^\/say (.+)$/, (msg, props) => {
  const text = props.match[1];
  console.log(msg);
  return bot.sendMessage(msg.chat.id, text, { replyToMessage: msg.message_id });
});

In the previous code we initialized the library that we will use to communicate with the telegram API as well as writing a very simple command so that when using /say anything the bot responds anything.

In order not to write directly the bot token in our code, we previously used the TG_TOKEN variable, so now we must create a .env file and install dotenv in our project. This module allows us to load environment variables from a .env file to process.env, hence instead of writing our 1778008574:AAFbuYsrLW3Q77... token in the bot.js file, we write process.env.TG_TOKEN.

Install dotenv

npm install dotenv

We create an .env file and write our variable and its value in it

TG_TOKEN=123456789:AAFbuKsrLW3Q77HsElI7oHGFqJXItozZ2jQ

There is only one simple step left to test our bot…

Let’s modify package.json to use dotenv. Let’s change the start script as follows:

"scripts": {
    "start": "node -r dotenv/config bot.js"
  },

Now yes, to start our bot, let’s type npm run start.

In a future post we will see how to host it in railway.app (currently it is running from our PC)