"May 14, 2016" Twitter search tool

This was created to support a collaborative project by Miranda July and Paul Ford with support from Starlee Kine as part of the Rhizome 7on7 project.

This is a Flask app and some utilities that load data into SQLite. Standard caveats about speed of development and general incompetence apply/but has the merit of working okay.

It makes it possible to search and browse historical tweets across a group of individuals. For example: Load in all the people in your Slack room, search their tweets for "dream," and recite their dreams back to them.

Up and running:

• python 3

• put all the twitter handles in data/handles.csv

• pip install -r requirements.txt

• go get a twitter API key

• copy config_template.py to config.py and modify per that API key

• Run these things

  • python -m util.make_db # make the database
  • python -m util.load_users # load the users from data/handles.csv
  • python -m util.import_tweets # import the tweets from those handles

Tweet importing will take a long time, many hours or days depending on your list of people, their prolixity, etc.

Run the app:

• python app.py

Notes

• Works fine under gunicorn.
• If you start to mess with itwhile the tweets are loading into SQLite it can skip tweets because SQLite wants none of that.