Markov text-generator for Twitter timelines. Deals with @ mentions, hashtags, and URLs.
With pip:
pip install twarkov
See example.py. You'll need Twitter API credentials to scrape a timeline, and to set them as environment variables for the example script to work. First replace the '???' below with your creds and run these commands in your terminal:
export APP_KEY='???'
export APP_SECRET='???'
export OAUTH_TOKEN='???'
export OAUTH_TOKEN_SECRET='???'
Then clone this repo, create a virtualenv, and install requirements:
git clone https://github.com/amandapickering/twarkov.git
mkvirtualenv twarkov
pip install -r requirements.txt
Edit example.py to add your twitter handle, then run it:
python example.py
You'll see a markov-generated tweet each time you run the script.
you can also tweak the ngram–example.py uses 1 for more randomness, 2 (the default) works as well.
- still some issues with apostrophes in the middle of words
- no conjunctions at the end of a tweet
- TESTS