Make ordinary tweets fleet-like with AWS Lambda and Python

What is a fleet?

Fleet provides the ability to take advantage of Twitter and communicate on Twitter in an unprecedented way. The content shared on Fleet disappears after 24 hours, so you can share your thoughts and feelings of depression. You can share your temporary personal thoughts with your followers without any reaction from other general users. Fleet creators can click on their Fleet and tap the read text at the bottom to see who viewed their Fleet, including accounts that keep tweets private.

--It disappears after 24 hours ――In other words, you only have to delete it once every 24 hours.

Fleet drawbacks

--There is no comment function and you will be notified by DM --Even if you send an emoji, you will be notified by DM ――I still want Fav, RT, and reply functions, just like regular tweets.

How to fleet tweets?

--Clarify "tweets to leave" and "tweets to disappear" --Conditions for tweets to be left --There is a Fav --There is RT --There is a reply --Conditions for erasing tweets --Everything except the tweets you leave --According to the above conditions, fire Lambda once every 24 hours and delete the tweet

Prerequisites

--Twitter API key issued ――I will do my best because English composition will come out as much as I can if I google --Created an AWS account --Keep AWS CLI available

Build

Folder structure

DeleteNoRtFavReplyTweets
|   lambda.yml
|   README.md
|
\---code
        DeleteNoRtFavReplyTweets.py

Cloudformation

AAWSTemplateFormatVersion: '2010-09-09'
# Twitter API Key
Parameters:
  ConsumerKey:
    Type: String
    NoEcho: true
  ConsumerSecret:
    Type: String
    NoEcho: true
  AccessKey:
    Type: String
    NoEcho: true
  AccessSecret:
    Type: String
    NoEcho: true
Resources:
# Lambda
  Lambda:
    Type: 'AWS::Lambda::Function'
    Properties:
      Code: code
      Environment: 
        Variables:
          ConsumerKeyVar: !Ref ConsumerKey
          ConsumerSecretVar: !Ref ConsumerSecret
          AccessKeyVar: !Ref AccessKey
          AccessSecretVar: !Ref AccessSecret
      Description: NoRtFavReplyTweetDelete
      FunctionName: NoRtFavReplyTweetDelete
      Handler: noRtFavTweetDelete.lambda_handler
      MemorySize: 128
      Role: !Sub 
        - arn:aws:iam::${AWS::AccountId}:role/${Domain}
        -  { Domain: !Ref LambdaRole }
      Runtime: python3.8
      Timeout: 180
  # CloudWatchEvents Rule
  Rule:
    Type: 'AWS::Events::Rule'
    Properties:
      Description: NoRtFavReplyTweetDeleteRule
      Name: NoRtFavReplyTweetDeleteRule
      ScheduleExpression: 'cron(0 17 * * ? *)' #2 o'clock at night
      State: ENABLED
      Targets:
        - Arn: !GetAtt Lambda.Arn
          Id: lambda
  # Lambda IAM Role
  LambdaRole:
    Type: 'AWS::IAM::Role'
    Properties:
      RoleName: deleteTweetRole
      AssumeRolePolicyDocument:
        Version: '2012-10-17'
        Statement:
          - Effect: Allow
            Principal:
              Service:
                - lambda.amazonaws.com
            Action:
              - 'sts:AssumeRole'
      Path: /
      ManagedPolicyArns:
        - arn:aws:iam::aws:policy/service-role/AWSLambdaBasicExecutionRole
  # CloudWatchEvents Lambda excute enable
  LambdaEvent:
    Type: 'AWS::Lambda::Permission'
    Properties:
      Action: 'lambda:InvokeFunction'
      FunctionName: !Ref Lambda
      Principal: 'events.amazonaws.com'
      SourceArn: !GetAtt Rule.Arn

Python

import tweepy
import os

# API Key
consumer_key = os.environ['ConsumerKeyVar']
consumer_secret = os.environ['ConsumerSecretVar']
access_key = os.environ['AccessKeyVar']
access_secret = os.environ['AccessSecretVar']

# Tweepy Auth
auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
auth.set_access_token(access_key, access_secret)
api = tweepy.API(auth, wait_on_rate_limit = True)

noFavRtTweet = []
mentionTweet = []
noDeleteTweet = []

def lambda_handler(event, context):
    # No Favorite No RT Tweet
    print("==========No Favorite No RT Tweet==========")
    for tweet in tweepy.Cursor(api.user_timeline,exclude_replies = True).items():
        if tweet.favorite_count == 0 and tweet.retweet_count == 0: #Change according to the number of Favs
            print(tweet.id,tweet.created_at,tweet.text.replace('\n',''))
            noFavRtTweet.append(tweet.id)

    # Reply Tweet
    print("==========Reply Tweet==========")
    for mentions in tweepy.Cursor(api.mentions_timeline).items():
        print(mentions.id,mentions.created_at,mentions.text.replace('\n',''))
        mentionTweet.append(mentions.in_reply_to_status_id)

    print("==========No Favorite No RT Reply Tweet==========")
    # No Favorite No RT Reply Tweet
    for tweet in noFavRtTweet:
        for reptw in mentionTweet:
            if tweet == reptw:
                print(api.get_status(tweet).id,api.get_status(tweet).created_at,api.get_status(tweet).text.replace('\n',''))
                noDeleteTweet.append(tweet)

    # Extraction Delete Tweet
    print("==========Extraction Delete tweet==========")
    perfectList = set(noFavRtTweet) ^ set(noDeleteTweet)
    print(list(perfectList))

    # Delete Tweet
    print("==========delete tweet==========")
    for deltw in perfectList:
        print(api.get_status(deltw).id,api.get_status(deltw).created_at,api.get_status(deltw).text)
        api.destroy_status(deltw)

Deploy to AWS

--Tweepy is an external module, so it is included.

pip install tweepy --target ./code

--Package and transfer to S3

aws cloudformation package --s3-bucket $YourBucketName `
--template-file lambda.yml `
--output-template-file lambda-packaged.yml

--Infrastructure construction and deployment --Here, specify the Twitter API KEY in the Lambda environment variable.

aws cloudformation deploy `
--template-file lambda-packaged.yml `
--stack-name $YourStackName `
--parameter-overrides AccessSecret=$YourAccessSecret `
ConsumerKey=$YourConsumerKey ` 
ConsumerSecret=$YourConsumerSecret ` 
AccessKey=$YourAccessKey `
--capabilities CAPABILITY_NAMED_IAM

--Check from the mannequin

Full version on GitHub

https://github.com/trysaildaisuki/DeleteNoReactionReply

truth

-This article is a by-product ――It is good to set countermeasures against Fabo explosion from the number of Favs of your tweets.