Serve data on lifecycle (open -> milestones -> close) of a repo's issues

Summarizes the events in a Github repository's issues. Output is available as

• An API (/api/<owner>/repo/), returning for each issue a list of spans - time periods defined by the milestones in effect during that time

• A Bokeh graph (/<owner>/<repo>/) illustrating each issue as a horizontal bar, with a different color for each span. The bars show milestone and assignee information when hovered over, and are hyperlinked to the issues' pages at Github.

To run the development server:

1. Create and activate a Python virtual environment

2. Install PostgreSQL (other SQLAlchemy-compatible databases are OK, too, if you modify config.py appropriately)

3. Run pip install -r requirements.txt

4. createdb github-issues-dev

5. Run python setup.py db upgrade to create the database tables.

6. Run python manage.py deploy

7. Visit http://0.0.0.0:5000//

If you want to report on a private repository, set the environment variables GITHUB_USER and GITHUB_AUTH. This will also protect your server for running up against Github's unauthenticated API usage limits.

The local server stores information queried from Github in a local PostgreSQL database, so that it will save web traffic and load on Github. The server will rely entirely on locally stored data if its last query to Github was more than REFRESH_THRESHHOLD_SECONDS ago, a constant stored in config.py (default 1 hour).