The world of Ian Hill

Twitterbots - Tweeting replies to mentions

In the previous pages we learned the basics on how to post tweets, pictures and shout-out to new followers to your Twitter page by way of a Twitter app

Pages in this series...

Page 1: Overview
Page 2: Tweeting randomly
Page 3: Tweeting pictures
Page 4: Tweeting new followers
Page 5: Tweeting replies to mentions
Page 6: Scanning for trending hashtags
Page 7: Automating everything

Previous code example

#!/usr/bin/env python
import sys
import os
import subprocess
import time
import datetime
from twython import Twython
from time import strftime

CONSUMER_KEY = 'YOUR CONSUMER KEY GOES HERE'
CONSUMER_SECRET = 'YOUR CONSUMER KEY SECRET GOES HERE'
ACCESS_KEY = 'YOUR ACCESS KEY GOES HERE'
ACCESS_SECRET = 'YOUR ACCESS SECRET GOES HERE'

twitter = Twython(CONSUMER_KEY, CONSUMER_SECRET, ACCESS_KEY, ACCESS_SECRET)

mentions = []
users = []
twitter_id = []
tweetID = []
ts = int(time.time())

try:
   tweets = twitter.get_mentions_timeline(count=100)
   for tweet in tweets:
      users.append("@" + str(tweet['user']['screen_name']) + "\n")
      tweetID.append(str(tweet['id_str']) + "\n")
      s = str(tweet['created_at']) # Thu Oct 26 13:11:02 +0000 2017
      t = time.mktime(datetime.datetime.strptime(s, "%a %b %d %H:%M:%S +0000 %Y").timetuple())
      ti = int(t)
      mentions.append(str(ti) + " # @" + tweet['user']['screen_name'] + " - " + tweet['text'].replace('\n', ' ') + " - " + tweet['id_str'] + "\n")
      tp = ts - ti
      twitter_id.append(tweet['id_str'])
except:
   print str(sys.exc_info())
   sys.exit()

for tweet in mentions:
   print tweet

If you run the code in a terminal session, it'll print a list of your last 100 mentions to the terminal, which is all lovely and that, but we should do something else with the code instead, it's just the way we are!

# New imports to be aware of
#!/usr/bin/env python
import sys
import os
import subprocess
import time
import datetime
from twython import Twython
from time import strftime

CONSUMER_KEY = 'YOUR CONSUMER KEY GOES HERE'
CONSUMER_SECRET = 'YOUR CONSUMER KEY SECRET GOES HERE'
ACCESS_KEY = 'YOUR ACCESS KEY GOES HERE'
ACCESS_SECRET = 'YOUR ACCESS SECRET GOES HERE'

twitter = Twython(CONSUMER_KEY, CONSUMER_SECRET, ACCESS_KEY, ACCESS_SECRET)

# Load previous tweet ID list into an array
f = open('/var/www/html/users.txt', 'r')
prev = f.readlines()
f.close()

# Set up variables used in the code
mentions = []
users = []
twitter_id = []
tweetID = []
try:
   tweets = twitter.get_mentions_timeline(count=100)
   for tweet in tweets:
      users.append("@" + str(tweet['user']['screen_name']) + "\n") # This generates the list of screen names
      tweetID.append(str(tweet['id_str']) + "\n") # This generates the list of tweet ID's
      s = str(tweet['created_at']) # Thu Oct 26 13:11:02 +0000 2017 # This gets the time of the mention
      t = time.mktime(datetime.datetime.strptime(s, "%a %b %d %H:%M:%S +0000 %Y").timetuple()) # This converts the formatted time back to a Unix timestamp
      ti = int(t)
      mentions.append(str(ti) + " # @" + tweet['user']['screen_name'] + " - " + tweet['text'].replace('\n', ' ') + " - " + tweet['id_str'] + "\n") # This generates the list of mentions with all the data
      tp = ts - ti
      if (tp < 10800): # This checks if the mention was in the last three hours. If it was, the tweet ID gets added to the third array
         twitter_id.append(tweet['id_str'])
except Exception as e:
   print "Failed! " + str(e)
   sys.exit()

# Two files are written here, one for the mentions data, and one for the user ID's
users_text = open("/var/www/html/users.txt","w+")
users_text.write("".join(tweetID))
users_text.close()
mentions_text = open("/var/www/html/mentions.txt","w+")
mentions_text.write("".join (mentions).encode ("utf-8"))
mentions_text.close()

# This generates the list of random replies and assigns one to the tweetA variable
def randTweet():
   global tweetA
   twitStart=[]
   twitStart.append("Are you talking to me?")
   twitStart.append("Did you just say something?")
   twitStart.append("Sorry, did you call me some time ago?")
   twitStart.append("Sorry, were you talking to me a while ago?")
   twitStart.append("I'm ignoring you")
   twitStart.append("I'm just going to pretend you didn't say that")
   shuffle(twitStart)
   twitStartLength = len (twitStart)
   twitBegin = (randint(0,twitStartLength-1))
   tweetA = twitStart[twitBegin]

# This compares the two lists for differences
diff=[]
id=[]
z = 0
for a in tweetID:
   c = 0
   for b in prev:
      if (str(a) == str(b)):
         c = 1
   if (c == 0):
      diff.append(users[z])
      id.append(twitter_id[z])
   z = z + 1

# This is where the tweet is sent for each new reply
y = 0
for a in diff:
   time.sleep(1)
   randTweet()
   tweetStr = a.replace("\n","")+" "+tweetA
   ts = int(time.time())
   st = datetime.datetime.fromtimestamp (ts).strftime ('%H:%M:%S %d-%m-%Y')
   fileName = datetime.datetime.fromtimestamp (ts).strftime ('%Y%m%d')
   text_file = open("/var/www/html/" + fileName + ".log", "a")
   text_file.write(st + " (mentions) " + tweetStr)
   try:
      twitter.update_status(status=tweetStr, in_reply_to_status_id=id[y]) # This sends a tweet in reply to the tweet ID given, you may wish to comment this out the first time you run the script
   except:
      print str(sys.exc_info())
   y = y + 1

This script will load the previous tweet ID list that's in the /var/www/html folder into an array, then download the last 100 mentions from Twitter into a second array, separate them out into users and tweet ID's, then overwrite the file with the previous tweet ID list with the updated tweet ID list for next time, then will compare the lists and any mentions in the new array which are not in the old array get added to a third array, which then gets iterated over to reply to the new mentions.

The script just sees the tweet ID to reply to and tweets a random reply to that tweet, there is no reading of the tweet to give a specific reply, I haven't coded that yet...

There is a check so any reactivated accounts don't get their tweets to Vader replied to again, this simply checks to see if the new mention was within the last three hours, and as this script runs once an hour, it seems to work well

Simple?

As before here's the code to write each tweet to a text file

ts = int(time.time())
st = datetime.datetime.fromtimestamp ts).strftime ('%H:%M:%S %d-%m-%Y')
fileName = datetime.datetime.fromtimestamp (ts).strftime ('%Y%m%d')
text_file = open("/var/www/html/" + fileName + ".log", "a")
text_file.write(st + " (mentions) " + tweet +"\n")
text_file.close()

Also you will need to add this command in somewhere to write the error in to the log if the tweet fails to be posted

text_file.write(" - " + str(sys.exc_info()) + "\n")

This is it for this part of the tutorial

Check out part six where we scan for trending hashtags

For more project ideas check out my other projects here

© 2024 IWH Software (Ian Hill)