Skip to main content

Python utilities for twitter bots

Project description

twitter bot utils

Twitter bot utils make it a little easier to set up a Twitter bot, with an eye to making config and command-line options easy to manage and reproduce. They're intended for managing a small-to-medium-sized coterie of Twitter accounts on one machine. The package is a super-simple wrapper for the excellent Tweepy library. It also provides shortcuts for setting up command line tools with argparse.

This package is intended to assist with the creation of bots for artistic or personal projects. Don't use it to spam or harrass people.

Works with python >= 3.6.

Install with pip install twitter_bot_utils.

See a basic run through in the Hello World section of the documentation.


One hurdle with setting up bots is getting the proper authentication keys. It can be a bit of a pain to log in and out of Twitter's app site. Twitter bot utils comes with tbu auth, a command line helper for this:

$ tbu auth --consumer-key 1233... --consumer-secret 345...

This will prompt you with an url. Open this in a browser where your bot is logged in, click "Authorize". Twitter will show you an authorization code, enter this on the command line, and presto! your keys will be displayed.

tbu auth is inspired by a feature of twurl, Twitter's full-fledged command line tool.

Config files

One goal of Twitter Bot Utils is to create Tweepy instances with authentication data stored in a simple config file. This gives botmakers a simple, reusable place to store keys outside of source control.

By default, Twitter bot utils looks for a file called bots.yaml (or .yml) in the current directory, your home directory (~/) or the ~/bots directory. Custom config locations can be set, too.

These are two ways to lay out a bots config file. The basic way covers just one user and one app:

token: ...
secret: ...
consumer_key: ...
consumer_secret: ...
my_setting: "bots are good"

If you have more than one bot, a simple setup is to have one app for each bot. Visit, register the app, and then choose "Create my access token" in the "keys and tokens" tab.

general_setting: "all bots share this setting"

    # twitter screen_name
        token: ...
        secret: ...
        consumer_key: ...
        consumer_secret: ...
        custom_setting: "bots are great"


If you have one app shared by several bots, create an apps section in the config file:

        consumer_key: ...
        consumer_secret: ...
        token: ...
        secret: ...
        app: my_app_name

The twitter-auth utility will happily read settings from a bots.yaml file set up like this:

twitter-auth -c ~/bots.yaml --app my_app_name

Using config files to talk to Twitter

Using a config file in one of the default locations doesn't require any extra settings:

import twitter_bot_utils as tbu

# Automatically check for a config file in the above-named directories
twitter = tbu.API(screen_name='MyBotName')

The twitter object is a fully-authenticated tweepy API object. So you can now do this:

twitter.update_status(status='hello world')

The bots config file is also useful for storing keys and parameters for other APIs, or for your own bots.

# Get a config settings from your bots config file. This might be the key for a third-party API
# Use a general setting
# "all bots share this setting"

# Settings from the user and app section are also available:
# "bots are great"

# "apple juice"

Set a custom config file with the config_file argument:

# Specify a specific config file
twitter = tbu.API(screen_name='MyBotName', config_file='path/to/config.yaml')

Twitter bot utils comes with some built-in command line parsers, and the API object will also happily consume the result of argparse.parser.parse_args() (see below for details).

Without user authentication

Some Twitter API queries don't require user authentication. To set up an Tweepy API instance without user authentication, set up a bots.yaml file as above, but omit the users section. Use the app keyword argument:

twitter = tbu.API(app='my_app_name', config_file='path/to/config.yaml')"Twitter searches don't require user authentication")

Recent tweets

The twitter_bot_utils.API object extends tweepy.API with some methods useful for bots:

  • Methods to check for the ID of recent tweets: last_tweet, last_reply, last_retweet. These are useful if your bot searches twitter and wants to avoid ingesting the same material.
twitter = tbu.API(screen_name='MyBotName')

# id of most recent tweet from MyBotName

# id of most recent reply from MyBotName

# id of most recent retweet from MyBotName

# Example: what's happened since the last time the bot was active?'#botALLY', since_id=twitter.last_tweet)

Twitter bot utils also adds a retry in update_status when Twitter is over capacity. If update_status gets a 503 error from Twitter, it will wait 10 seconds and try again.

Default Command Line Options

It's useful to package bots as command line apps so that they can be easily run with cron. Twitter bot utils includes some helpers for working with argparse.

Some useful command line flags are available by default:

  • -u, --user: Screen name to run as
  • -n, --dry-run: Don't tweet, just output to stdout
  • -v, --verbose: Log to stdout
  • -q, --quiet: Only log errors
  • -c, --config: path to a config file. This is a JSON or YAML file laid out according to the above format. This option isn't needed if the config file is in one of the default places.

Say this is

import argparse
import twitter_bot_utils as tbu

# This sets up an argparse.ArgumentParser with the default arguments
parent = tbu.args.parent()
parser = argparse.ArgumentParser('My Example Bot', parents=[parent])
parser.add_argument('--my-arg', type=str, help='A custom argument')

args = parser.parse_args()

# Set up the tweepy API
# Note that you can pass the argparse.Namespace object
twitter = tbu.API(args)

# Generate a tweet somehow
tweet = my_tweet_function(args.my_arg)

# The API includes an instance of logging
# debug logs will output to stdout only if --verbose is set
# info logs will output even without --verbose
api.logger.debug("Generated %s", tweet)

# Use args.dry_run to control tweeting
if not args.dry_run:

Then on the command line:

> python --help
usage: [options]

My Example Bot

optional arguments:
  -h, --help            show this help message and exit
  -c PATH, --config PATH
                        bots config file (json or yaml)
                        Twitter screen name
  -n, --dry-run         Don't actually do anything
  -v, --verbose         Run talkatively
  -q, --quiet           Run quietly
  --my-arg MY_ARG       A custom argument

# Looks for settings in a config file (e.g. bots.yaml, see config section above)
# Prints results to stdout and doesn't publish anything 
> python  --dry-run --verbose

# Run quietly, say in a crontab file
> python --user MyBotName --quiet
Generated <EXAMPLE TWEET 2>


Checking for entities

Easily check if tweets have specific entities:

import twitter_bot_utils

# Don't set include_entities to False and expect the below to work
statuses ='example search', include_entities=True)

status = status[0]

# returns True if status has one or more mentions, otherwise False 

# returns True if status has one or more hashtags, otherwise False 

# returns True if status has one or more media entities (images, video), otherwise False 

# returns True if status has any entities

# These also exist:

Filtering out entities

These helpers remove entities from a tweet's text.

import twitter_bot_utils as tbu

api = tbu.API(screen_name='MyBotName')

results ="special topic")

# 'This is an example tweet with a #hashtag and a link'

tbu.helpers.remove_entity(results[0], 'hashtags')
# 'This is an example tweet with a  and a link'

tbu.helpers.remove_entity(results[0], 'urls')
# 'This is an example tweet with a #hashtag and a link '

# Remove multiple entities with remove_entities.
tbu.helpers.remove_entities(results[0], ['urls', 'hashtags', 'media'])
# 'This is an example tweet with a  and a link '

Command line utilities

Twitter bot utils includes a command line tool with a few useful subcommands:

  • tbu auth: Authenticate and account with a Twitter app.
  • tbu follow: Follow accounts that follow your bot
  • tbu like: Like (aka favorite) your bot's mentions
  • tbu post: Basic command line for posting text and images

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

twitter_bot_utils-0.14.0.tar.gz (48.8 kB view hashes)

Uploaded Source

Built Distribution

twitter_bot_utils-0.14.0-py2.py3-none-any.whl (30.7 kB view hashes)

Uploaded Python 2 Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page