Get celebrity tweet history from twitter

A wall that I immediately ran into when I wanted to use IBM cloud's personality insight. A memorandum at that time.

trouble

A sentence of 3000 words or more is required (or rather desirable). So I borrowed a celebrity tweet from twitter. It is considered that there are about 15 words per tweet, and 200 tweets are acquired per person.

However, as a major premise, it is necessary to register the twitter API. I have already applied for registration, so I will skip this step here.

Implementation method

 -*- coding:utf-8 -*-

import tweepy
import re
import subprocess

# User list
import user_list
# twitter API authentication key
# Access_token, Access_secret, Consumer_key, Consumer_secret
from auth import twitter_credentials as tc


def get_twitterdata(username, rfile):

 #Authentication key reading, API setting
  auth = tweepy.OAuthHandler(tc.Consumer_key, tc.Consumer_secret)
  auth.set_access_token(tc.Access_token, tc.Access_secret)
  api = tweepy.API(auth, wait_on_rate_limit = True)

 #List to store tweets
  tweets_data =[]
 Data acquisition up to # 200 tweet
  for tweet in api.user_timeline(screen_name=username, count=200):
 # Get tweet text
    tmp_text=tweet.text
 #Continuous line breaks are combined into one
    tmp_text=re.sub('\n+','\n',tmp_text)
 #Add tweet to list
    tweets_data.append(tmp_text + '\n')

 # File output
  with open(rfile, "w",encoding="utf-8") as wf:
    wf.writelines(tweets_data)


if __name__ == '__main__':

 Get your #twitter username
  userlist=user_list.username

  for i in range(0,len(userlist)):
    username = userlist[i]
    rfile = "./data/tweet_"+str(i).zfill(3)+".csv"

    try:
      get_twitterdata(username, rfile)
 # Generate an empty file if it cannot be obtained, such as when it is set to private
    except:
      subprocess.run(["touch",rfile])

User list example

It's crazy, but the person's name is hidden. username=[ "ariyoshihiroiki", "matsu_bouzu", "takapon_jp" ]

result

The description is omitted because it may be copyrighted. tweet_000.csv tweet_001.csv tweet_002.csv Can be done.

Afterword

Since the amount of data that can be acquired every 15 minutes (?) Is limited, if you are greedy, you will have to wait a long time. You may also exclude retweets and other non-texts from the person in question.