typescript-telegram-bot-api

📦 typescript-telegram-bot-api

This is a TypeScript wrapper for the Telegram Bot API Node.js and browsers. It allows you to easily interact with the Telegram Bot API using TypeScript. Check out the browser live demo here: StarExplorer.

Installation

npm install typescript-telegram-bot-api

Usage

import { TelegramBot } from 'typescript-telegram-bot-api';

const bot = new TelegramBot({ botToken: 'YOUR_BOT_TOKEN' });
bot.startPolling();

bot.on('message', (message) => {
  console.log('Received message:', message.text);
});

bot.on('message:sticker', (message) => {
  console.log('Received sticker:', message.sticker.emoji);
});

bot.getMe()
  .then(console.log)
  .catch(console.error);

Supported Methods

Descriptions of methods can be found at https://core.telegram.org/bots/api#available-methods. Method names and parameters correspond to those in the official API documentation. Each method returns a Promise that resolves with data received from the API.

In case of an API error, the Promise will be rejected with a TelegramError containing the error code and description from the API. If the API error includes a retry_after field, the library will retry the request after the specified number of seconds, until a response without an error is received. This behavior can be disabled by setting the autoRetry parameter to false.

If the error is not related to the API, the Promise will be rejected with a different error.

For sending files, you can use not only 'file_id' or 'url', but also stream.Readable or Buffer. To send files with additional parameters, such as a filename or specific contentType, use the FileOptions wrapper class.

import { TelegramBot, FileOptions } from 'typescript-telegram-bot-api';
import { createReadStream } from 'fs';
import { readFile } from 'fs/promises';


await bot.sendPhoto({
  chat_id: chat_id,
  photo: 'AgACAgIAAxkDAAIF62Zq43...AgADcwADNQQ',
  caption: 'file_id',
});

// or

await bot.sendPhoto({
  chat_id: chat_id,
  photo: 'https://unsplash.it/640/480',
  caption: 'url',
});

// or

await bot.sendPhoto({
  chat_id: chat_id,
  photo: createReadStream('photo.jpg'),
  caption: 'stream',
});

// or 

await bot.sendPhoto({
  chat_id: chat_id,
  photo: await readFile('photo.jpg'),
  caption: 'buffer',
});

// or 

await bot.sendPhoto({
  chat_id: chat_id,
  photo: new FileOptions(
    await readFile('photo.jpg'), {
      filename: 'custom_file_name.jpg',
      contentType: 'image/jpeg',
    }
  ),
  caption: 'FileOptions',
});

// or in browser

await bot.sendPhoto({
  chat_id: chat_id,
  photo: input.files[0], // or new File(…)
  caption: 'file',
});

Events

TelegramBot is an EventEmitter that emits the Update event and also emits events for each type of Message, such as message:audio, when the audio field is present in the message object.

bot.on('message', (message) => {
  console.log('Received message:', message.text);
});

bot.on('message_reaction', (messageReactionUpdated) => {
  console.log('Received message_reaction:', messageReactionUpdated);
});

bot.on('message:audio', (message) => {
  console.log('Received audio:', message.audio.file_id);
});

Error Handling

Wrap asynchronous calls in try...catch blocks or use .catch() on promises to handle exceptions properly.

try {
  await bot.sendPhoto({
    chat_id: chat_id,
    photo: createReadStream('photo.jpg'),
    caption: 'stream',
  });
} catch (error: unknown) {
  if (TelegramBot.isTelegramError(error)) {
    // Handle Telegram API errors
    if(error.response.description === 'Bad Request: chat not found'){
      console.info('Message not sent: chat not found');
    } else if (error.response.description === 'Request Entity Too Large'){
      console.info('Message not sent: file too large');
    } else if (error.response.description === 'Bad Request: IMAGE_PROCESS_FAILED'){
      console.info('Message not sent: image processing failed');
    } else if (error.response.description === 'Forbidden: bot was blocked by the user'){
      console.info('Message not sent: user blocked bot');
    } else {
      console.error('Telegram API Error:', error.message);
    }
  } else if (error instanceof Error) {
    // Handle system errors (example: no such file)
    console.error('System Error:', error.message);
  } else {
    console.error('Unknown Error:', error);
  }
}

Webhooks

To use webhooks, you need to set up a server that will receive updates from Telegram. You can use the express library for this purpose.

This example demonstrates a basic Telegram bot that responds with a reaction to any incoming message using Webhooks. The use of ngrok as a tunneling service simplifies the setup, allowing the bot to be easily deployed in various environments without complex network configuration. This approach is ideal for quick testing and development purposes. For production use, you should consider deploying the bot on a server with a public IP address and a valid SSL certificate.

import 'dotenv/config';
import * as ngrok from 'ngrok';
import express from "express";
import {TelegramBot} from "./src";
import {Update} from "./src/types";

const port = 3001;

const bot = new TelegramBot({
  botToken: process.env.TEST_TELEGRAM_TOKEN as string,
});

bot.on('message', async (message) => {
  await bot.setMessageReaction({
    chat_id: message.chat.id,
    message_id: message.message_id,
    reaction: [{
      type: 'emoji', emoji: '👍'
    }]
  });
});

const app = express();
app.use(express.json());
app.post('/', async (req, res) => {
  try {
    await bot.processUpdate(req.body as Update);
    res.sendStatus(200);
  } catch (e) {
    res.sendStatus(500);
  }
});

(async () => {
  app.listen(port, async () => {
    const url = await ngrok.connect({
      proto: 'http',
      addr: port,
    });
    await bot.setWebhook({url});
    console.log('Set Webhook to', url);
  })
})();

process.on('SIGINT', async () => {
  await bot.deleteWebhook();
  await ngrok.disconnect();
  console.log('Webhook deleted');
});

Tests

npm test

docker build -t typescript-bot-api .
docker run --rm --env-file .env typescript-bot-api run test

CI/CD is set up with GitHub Actions. Tests and linters are run on every pull request.

If you want to run tests locally, follow the instructions in tests/README.md.