spindle is the web front-end for the loom graph processing stack. It can be used to search, browse and edit entities that are stored in a loom-generated statement database.

It is implemented as an AngularJS-based single page application, which requests data from a Flask-based backend service. Assets are managed via bower and compiled using webassets.

Please read the loom introduction before installing or developing spindle.

For spindle to run, you need to first make sure to have a loom statement database and ElasticSearch index prepared and for these services to be accessible from spindle.

You will also need to install the following dependencies:

• sass (usually installed via ruby gems, ruby-sass on Debian)
• node and npm are needed to install both bower and uglifyjs

It is recommended to run spindle inside a virtual environment to separate it's dependencies from the system Python. In that case, you can install spindle and it's dependencies like this:

# If you haven't yet:
$ sudo npm install -g bower uglifyjs
$ sudo apt-get install ruby-sass
# Then, the application itself:
$ git clone git@github.com:occrp/spindle.git
$ cd spindle
$ virtualenv pyenv
$ source pyenv/bin/activate
$ pip install -r requirements.txt
$ pip install -e .
$ bower install

You also need to create configuration file to define local settings, such as database access and the location of the ElasticSearch index:

$ cp settings.tmpl.py settings.py
# Now edit the settings.py and replace any configuration variables to match
# your local environment.

Finally, you can run spindle:

# If you open a new console, make sure to enter the virtual environment:
$ source pyenv/bin/activate
# Set the path of the configuration file:
$ export SPINDLE_SETTINGS=`pwd`/settings.py
# Run the server:
$ spindle runserver

When using spindle in production, consider using a proper WSGI web server such as gunicorn instead of the built-in development server. A Dockerfile is provided for containerized deployment.

In order to run the tests, you need to have an instance of ElasticSearch running and configured in the current spindle configuration. The tests will create their own index (spindle_test_idx_) and destroy it automatically upon teardown.

Run tests like this:

$ pip install nose coverage
$ nosetests --with-coverage --cover-package=spindle --cover-erase

spindle is free software; it is distributed under the terms of the Affero General Public License, version 3.