BoraxTheClean/TwitterBot

A basic Twitter Bot on AWS Lambda.

TwitterBot

Setup

Apply for a developer account at https://developer.twitter.com/

We need to get four secrets from Twitter to use their API:

• ACCESS_SECRET
• ACCESS_TOKEN
• CONSUMER_KEY
• CONSUMER_SECRET

After you make an app you can get your CONSUMER_KEY and CONSUMER_SECRET from the Twitter API console.

Then you can generate your ACCESS_SECRET and ACCESS_TOKEN in the API console as well.

Confirm SSM Parameters

It is assumed that there are SSM parameters in your deployment region at CONSUMER_SECRET CONSUMER_KEY ACCESS_TOKEN ACCESS_SECRET. For help making SSM parameters: https://docs.aws.amazon.com/systems-manager/latest/userguide/sysman-paramstore-console.html

Lambda Configuration

Our Lambda Function has a name, twitter-bot, a handler, the method handler(event,context) in the file text.py, and we want to use the latest python3.8 runtime.

SimpleTwitterApp:
    Type: AWS::Serverless::Function
    Properties:
        FunctionName: twitter-bot
        CodeUri: dist/
        Handler: text.handler
        Runtime: python3.8

We give our Function basic SSM permisions, so it can fetch the parameters we configured for it.

Policies:
  - Version: '2012-10-17'
    Statement:
      - Effect: Allow
        Action: ssm:GetParameter
        Resource: '*'

Finally, we add an event to the function, sending out or tweet every 24 hours.

If you want to send tweets more often, you'll have to make them dynamic, twitter blocks repeat tweets in a 24 hour period.

Events:
  TweetEveryDay:
    Type: Schedule
    Properties:
      Schedule: rate(1 day)

Python Code

We use the tweepy sdk to access the twitter api and boto3 API to access SSM Parameter Store.

import tweepy
import boto3

'''
    AWS Client Initialization
'''

ssm = boto3.client('ssm')

Confirm SSM Parameters

Next we pull parameters from SSM using the AWS API. I use the WithDecryption flag to decrypt the values, since I chose the SecureString option when creating my parameters.

Lambda Configuration

Our Lambda Function has a name, twitter-bot, a handler, the method handler(event,context) in the file text.py, and we want to use the latest python3.8 runtime.

SimpleTwitterApp:
    Type: AWS::Serverless::Function
    Properties:
        FunctionName: twitter-bot
        CodeUri: dist/
        Handler: text.handler
        Runtime: python3.8

We give our Function basic SSM permisions, so it can fetch the parameters we configured for it.

Policies:
  - Version: '2012-10-17'
    Statement:
      - Effect: Allow
        Action: ssm:GetParameter
        Resource: '*'

Finally, we add an event to the function, sending out or tweet every 24 hours.

If you want to send tweets more often, you'll have to make them dynamic, twitter blocks repeat tweets in a 24 hour period.

Events:
  TweetEveryDay:
    Type: Schedule
    Properties:
      Schedule: rate(1 day)

Python Code

We use the tweepy sdk to access the twitter api and boto3 API to access SSM Parameter Store.

import tweepy
import boto3

'''
    AWS Client Initialization
'''

ssm = boto3.client('ssm')

Next we pull parameters from SSM using the AWS API. I use the WithDecryption flag to decrypt the values, since I chose the SecureString option when creating my parameters.

'''
    Twitter API Initialization
'''

try:
    CONSUMER_KEY = ssm.get_parameter(Name='CONSUMER_KEY',WithDecryption=True)['Parameter']['Value']
    CONSUMER_SECRET = ssm.get_parameter(Name='CONSUMER_SECRET',WithDecryption=True)['Parameter']['Value']
    ACCESS_TOKEN = ssm.get_parameter(Name='ACCESS_TOKEN',WithDecryption=True)['Parameter']['Value']
    ACCESS_SECRET = ssm.get_parameter(Name='ACCESS_SECRET',WithDecryption=True)['Parameter']['Value']
except Exception as error:
    print(error)
    print("Failed to fetch api keys.")
    raise error

Then we initialize our tweepy client with the decrypted secrets we pulled in the previous steps.

auth = tweepy.OAuthHandler(CONSUMER_KEY, CONSUMER_SECRET)

auth.set_access_token(ACCESS_TOKEN, ACCESS_SECRET)

api = tweepy.API(auth)
user = api.me()
print('Tweeting for '+user.name)

Finally, we define a handler for Lambda to invoke our function through. We wrap our business logic in convenient functions, and we're done.

def handler(event, context):
    text = make_tweet()
    send_tweet(text)

def make_tweet():
    return "I've completed my Quest!\n\nI built this Lambda Bot using @AWSOwen's course!"


def send_tweet(text):
    print(text)
    api.update_status(text)

Deployment

Modify this line of script.sh and replace owen-lambda-bucket with your own s3 bucket located in the region you are deploying into.

BUCKET=owen-lambda-bucket

Then simply run

make deploy

Deploy and Enjoy

Have fun coding! And be sure to follow me on GumRoad for new courses and updates!

Find me on Twitter.

Shoot me an email.