Need help with this tutorial? Join the Botmakers community!
Update #2¶
Twitter discontinued the API this tutorial relied on and introduced webhooks instead. Check out my Glitch starter project that shows how to use them.
Update¶
See this project that shows how to properly handle responding to @ mentions and DMs on Glitch.
Due to how Twitter API works, Glitch is not a suitable platform for hosting this kind of a bot. Feel free to follow the tutorial below to learn about how Twitter bots work. Alternatively, you can take your finished code and find a hosting platform that lets you run your app non-stop.
Introduction¶
Hey there, Stefan here, back with another Twitter bot tutorial.
If this is your first time on Botwiki, be sure to read a bit about bots to get an idea what they are, and what kind of bots you can make.
In this tutorial, I am going to use the amazing Glitch website (formerly known as Gomix, even more formerly known as HyperDev), which lets you create and remix all kinds of fun apps, including bots.
Also, we are going to write our bot with node.js and the Twit library, but you don’t need to install anything on your computer — all the magic will happen inside Glitch.
If you’d like, you can have a quick look the finished code before we start.
Sounds good? Let’s do this!
Everything is a remix¶
First, you will need to create a new Twitter app; each comes with a set of free API keys which we will need to bring our Twitter bot to life.
Once you take care of that, go ahead and remix the Glitch Twitter bot template.
You can put your API keys inside your .env file (look for it in the sidebar), so that it looks something like this:
CONSUMER_KEY='1234gibberish'
CONSUMER_SECRET='54321somemoregibberish'
ACCESS_TOKEN='123secret-456stuff'
ACCESS_TOKEN_SECRET='987evenmoresecretstuff'
BOT_USERNAME='username'
Make sure to also update BOT_USERNAME to your bot’s actual username, without the @ sign.
To make our code easier to manage — and remix in the future, we are going to create separate functions, each handling a part of the bot’s behavior:
get_tweet_id: this function will extract the ID of a Tweet from the Tweet’s URLsend_twitter_dm: this will let the bot communicate with usretweet_tweet: this function will retweet a Tweet by given ID- and finally, we will put everything together with
user_stream.on('direct_message')
Before we begin, we need to slightly modify the first few lines of the server.js file.
var path = require('path'),
request = require('request'),
Twit = require('twit'),
config = {
/* Be sure to update the .env file with your API keys. See how to get them: https://botwiki.org/tutorials/make-an-image-posting-twitter-bot/#creating-a-twitter-app*/
twitter: {
consumer_key: process.env.CONSUMER_KEY,
consumer_secret: process.env.CONSUMER_SECRET,
access_token: process.env.ACCESS_TOKEN,
access_token_secret: process.env.ACCESS_TOKEN_SECRET
}
},
T = new Twit(config.twitter),
user_stream = T.stream('user');
Notice that I added one extra module called request, which will be used to figure out the ID of the tweet shared via DM. I am also listening to the user stream, instead of reading a sample of status updates (statuses/sample).
Getting the tweet ID¶
When you share a tweet via DM, you are basically just sending the tweet’s URL. On Twitter, all URLs are shortened in the format of https://t.co/abcd1234. We need to figure out the original URL (something like https://twitter.com/username/status/123456789), extract the ID, and then find a tweet with that ID and retweet it.
Our first function is going to look like this:
function get_tweet_id(url){
if (url.indexOf('https://twitter.com/')){
return url.split('/')[5];
}
else{
return false;
}
}
Sending a message¶
Next, we will give our bot the ability to speak by adding a generic function for sending direct messages. We can then use this function to respond with confirmation messages, or error messages if something breaks.
function send_twitter_dm(username, message){
T.post('direct_messages/new', {
screen_name: username,
text: message
},
function(err, data, response){
if (err){
console.log('Error!');
console.log(err);
}
});
}
Retweeting¶
And now, the retweeting function.
function retweet_tweet(tweet_id, user_name){
T.post('statuses/retweet/:id',
{
id: tweet_id
}, function(err, data, response) {
if (err){
send_twitter_dm(user_name, 'There was an error!\n\n' + err.message);
}
else{
send_twitter_dm(user_name, 'Retweeted!');
}
}
);
}
Pretty straightforward.
Our function will accept the ID of the tweet, which we will extract using the get_tweet_id function, and also the user_name of the person sharing the tweet, so we can either confirm that the tweet was successfully retweeted, or respond with an error message.
Tying it all together¶
We are going to listen to User Streams, which is a stream of data and events specific to the authenticated user.
In our case, we are going to wait for someone to message our bot.
Now, we don’t want to simply retweet every tweet someone shares with the bot. There’s a few things we could do to prevent someone hijacking our bot to post spam.
We could, for example, check if the bot is following the sender of the DM. Or we could create a Twitter list and check if the user is on that list. In this tutorial, I am going to go with the first option.
user_stream.on('direct_message', function (dm) {
T.get('friends/ids', { screen_name: process.env.BOT_USERNAME, stringify_ids: true }, function (err, data, response) {
(function(dm){
if (data.ids && data.ids.indexOf(dm.direct_message.sender.id_str) > -1){
var dm_text = dm.direct_message.text,
dm_sender = dm.direct_message.sender.screen_name;
if (dm_text.indexOf('https://t.co/') > -1){
/* Checking if the message is a shortened URL. */
var r = request.get(dm.direct_message.text, function (err, res, body) {
/* If yes, we will use the request module to get the original URL. */
if (r !== undefined){
var original_url = r.uri.href;
var tweet_id = get_tweet_id(original_url);
if (tweet_id){
retweet_tweet(tweet_id, dm_sender);
}
else{
send_twitter_dm(dm_sender, "This doesn't look like a tweet to me.");
}
}
else{
send_twitter_dm(dm_sender, "Couldn't get the original URL.");
}
});
}
else{
send_twitter_dm(dm_sender, "This doesn't look like a tweet to me.");
}
}
})(dm);
});
});
The code above might seem a bit more complicated, if you’re not familiar with closures in JavaScript, but since this is out of the scope of this tutorial, feel free to just copy and paste it as it is.
The bottom line is, we need to first fetch the list of accounts the bot is following, and then check if the sender of the DM is on that list. This is done asynchronously, meaning, the app won’t just wait while we load the list, and without going into too much detail, a closure is a way to deal with asynchronous events in JavaScript.
Sharing tweets¶
Moving on, we are now ready to test our bot. Make sure the bot follows you, then try sharing a tweet through a DM —
— and the tweet will be retweeted by your bot. If you don’t see the retweet, check out the finished code and make sure you’re not missing anything, and that you correctly updated your .env file.
There you go, that’s pretty much all there is to botmaking. Was it fun? Then be sure to share your creations with the Botmakers community!
Thanks for following along, and happy botmaking 🙂
