♌️ ⛹🏼 👲🏻 Writing a telegram bot on Go and AWS Lambda 👆🏼 👩🏿‍🚀 👦🏾

What do we do?

We will write a simple telegram bot that responds with the same message that was sent.

Why AWS Lambda?

Convenience of deployment, just write sls deploy

, and lambda is already unloaded
You only pay for the time lambda is running
No need to configure any servers and worry about scaling

What do you need?

Installed go
Nodejs and npm to install serverless
AWS account for deployment

TLDR

Clone the repository https://github.com/R11baka/echotgbot
Set a variable in the .env file BOT_TOKEN
Compiling the binary env GOOS=linux go build -o bin/webhook main.go
Unloading the lambda with sls deploy
Installing webhook with BOT_TOKEN

Sign up with AWS

AWS aws console aws_access_key_id, aws_secret_access_key .aws/credentials

.aws/credentials

cat ~/.aws/credentials
[default]
aws_access_key_id = ADEFEFEFFEBDXK3
aws_secret_access_key = Zy6ewfir/zGaT1B2/o9JDWDSssdrla
region = us-west-1

, BotFather. , BotFather /newbot

, , . , BotFather . , .

Serverless

Serverless- framework, , AWS Lambda . node, nodejs npm. serverless npm

npm install -g serverless

serverless ,

sls -v
Framework Core: 2.35.0 (standalone)
Plugin: 4.5.3
SDK: 4.2.2
Components: 3.8.2

serverless. serverless serverless.yml ,

service: echoBot
useDotenv: true
configValidationMode: error #   ,  , 
frameworkVersion: '>=1.28.0 <2.50.0'

provider:
  region: "us-west-1"
  lambdaHashingVersion: "20201221"
  name: aws
  runtime: go1.x
  logRetentionInDays: 30 #    
  endpointType: regional
  tracing: #    
    apiGateway: true
    lambda: true
  iam:
    role:
      statements:
        - Effect: "Allow"
          Resource: "*"
          Action:
            - "xray:*"
package:
  patterns:
    - "bin/webhook" #   

functions:
  webhook: 
    handler: bin/webhook
    timeout: 15
    description: simple echo bot
    memorySize: 128 #      
    environment:
      BOT_TOKEN: ${env:BOT_TOKEN}
    events:
      - http:
          path: /webhook
          method: ANY
          cors: false

Go

telebot.v2 aws-lambda-go

➜  go mod init testBot
go: creating new go.mod: module testBot
➜  go get -u gopkg.in/tucnak/telebot.v2
go: gopkg.in/tucnak/telebot.v2 upgrade => v2.3.5
go: github.com/pkg/errors upgrade => v0.9.1
➜ go get github.com/aws/aws-lambda-go
go: github.com/aws/aws-lambda-go upgrade => v1.23.0

2. main.go

package main

import (
	"encoding/json"
	"fmt"
	"github.com/aws/aws-lambda-go/events"
	"github.com/aws/aws-lambda-go/lambda"
	tb "gopkg.in/tucnak/telebot.v2"
	"os"
)

func main() {
	settings := tb.Settings{
		Token:       os.Getenv("BOT_TOKEN"),
		Synchronous: true,
		Verbose:     true,
	}
	tgBot, err := tb.NewBot(settings)
	if err != nil {
		fmt.Println(err)
		panic("can't create bot")
	}
	tgBot.Handle(tb.OnText, func(m *tb.Message) {
		message := m.Text
		tgBot.Send(m.Sender, message)
	})
	lambda.Start(func(req events.APIGatewayProxyRequest) (events.APIGatewayProxyResponse, error) {
		var u tb.Update
		if err = json.Unmarshal([]byte(req.Body), &u); err == nil {
			tgBot.ProcessUpdate(u)
		}
		return events.APIGatewayProxyResponse{Body: "ok", StatusCode: 200}, nil

	})
}

tucnak/telebot.v2 , return return events.APIGatewayProxyResponse{Body: "ok", StatusCode: 200}, nil

, .

Deploy

.env API_TOKEN BotFather

echo API_TOKEN={API_TOKEN_FROM_BOTFATHER} > .env

serverlss . .

sls print

env GOOS=linux go build  -o bin/webhook main.go

serverless

serverless deploy  -v

Service Information
service: echoBot
stage: dev
region: us-west-1
stack: echoBot-dev
resources: 11
api keys:
None
endpoints:
ANY - https://y7p31bwnu1.execute-api.us-west-1.amazonaws.com/dev/webhook
functions:
webhook: echoBot-dev-webhook
layers:
None

https://y7p31bwnu1.execute-api.us-west-1.amazonaws.com/dev/webhook => token , webhook

Integration with telegram

It remains to tell the telegram which endpoint to pull when receiving a message. This is done by the setWebhook command

curl https://api.telegram.org/bot{YOUR_TOKEN}/setWebhook?url={YOUR_DEPLOYED_AWS_URL}

Checking that the webhook is installed is done using getWebhookInfo

➜  ~ curl https://api.telegram.org/bot1324913549:AAE1zYMH6K3hF2TOgUQoIP-E1g4rMIamck/setWebhook\?url\= https://y7p31bwnu1.execute-api.us-west-1.amazonaws.com/dev/webhook
{"ok":true,"result":true,"description":"Webhook was set"}
➜  ~ curl https://api.telegram.org/bot1324913549:AAE1zYMH6K3hF2TOgUQoIP-E1g4rMIamck/getWebhookInfo
{"ok":true,"result":{"url":"https://y7p31bwnu1.execute-api.us-west-1.amazonaws.com/dev/webhook","has_custom_certificate":false,"pending_update_count":0,"max_connections":40,"ip_address":"184.169.148.254"}}

Errors

If something went wrong, go to CloudWatch and look at the logs, or you can also view the logs from the console

sls logs -f webhook

Writing a telegram bot on Go and AWS Lambda