DEV Community

Alba Rincón
Alba Rincón

Posted on

Building a Telegram bot: the bot comes to life

I recently got curious about how to develop a Telegram bot. That's why I have decided to give it a try and document the process in this series of posts, so anyone can benefit from it as well.

What are we building?👷‍♀️

The idea is really simple: we are building a dictionary bot. It will allow us to search for a word and give us back info about it like meanings, synonyms, etc. We want it to work in any conversation or group we have, so we can share it with others.
We will develop it in Golang using the WordsAPI and deploying it to Heroku.

Creating the bot in Telegram 🤖

To create a bot in Telegram we need to use the @BotFather bot.
We need to use the /newbot command, give it a name and a username and we'll get a link and a token in return:
Alt Text

We can use the link to start a conversation with our bot and later we'll also need the token to connect to the bot from our code. We can always go back to the @BotFather and retrieve again the token or change some settings.

Make it talk 🗣

We are ready to start the implementation. We'll be using the Telebot Go library. Telebot is a framework that is going help us a lot interacting with the Telegram Bot API (setting up the bot, handling requests, sending messages, etc).

Our only goal now will be making our bot reply to any message we send to it.

First of all, we need to create our bot instance and configure it with an endpoint that Telegram will use to POST the messages the bot receives.
That endpoint will be something like

https://localhost:[PORT]/bot[BOT_TOKEN]
Enter fullscreen mode Exit fullscreen mode

but it can't be in our local, we need to make it accessible to the internet so the bot can actually hit it.
For that purpose, we can use a tool called Ngrok and with a simple command it will expose the local port of our choice (4040 in this case) and give us back a URL.

ngrok http 4040
// returns a URL like https://89e834cae2b6.ngrok.io
Enter fullscreen mode Exit fullscreen mode

We have everything ready, so we can now create our bot instance configured with the webhook endpoint in which it will receive all the updates from Telegram:

webhook := &telebot.Webhook{
    Listen: ":" + os.Getenv("PORT"),
    Endpoint: &telebot.WebhookEndpoint{
        PublicURL: os.Getenv("WEBHOOK_URL") + "/bot" + os.Getenv("BOT_TOKEN"),
    },
}

settings := telebot.Settings{
    Token:  os.Getenv("BOT_TOKEN"),
    Poller: webhook,
}

bot, err := telebot.NewBot(settings)
if err != nil {
    log.Fatalf("error initializing bot: %s", err)
    return
}
Enter fullscreen mode Exit fullscreen mode

After that, we can add a handle to the bot that will trigger every time it receives any text message and replies to the user with I'm a 🤖!.

bot.Handle(telebot.OnText, func(msg *telebot.Message) {
    _, err := bot.Send(msg.Sender, "I'm a 🤖!")
    if err != nil {
        log.Printf("error sending message: %s", err)
        return
    }
})

bot.Start()
Enter fullscreen mode Exit fullscreen mode

We are finally ready to test our bot! Let's run it with:

PORT=[port of our chice] \
WEBHOOK_URL=[public url] \
BOT_TOKEN=[bot token provided by @BotFather] \
    go run cmd/dictiobot/main.go
Enter fullscreen mode Exit fullscreen mode

We can access the link that the @BotFather provided when we created the bot and send a message to it. Hopefully, we'll get our response back! 😅
Alt TextAnd it works! 🙌 It's not too clever for now, but it will get better!

You can check the whole code here: https://github.com/albarin/dictiobot/tree/01-the-bot-comes-to-life

In the next chapter, we will build a service to make requests to the WordsAPI to be able to search for words.

See you! 👋

Top comments (0)