The EULA Automated Analysis Tool is a web application that retrieves, parses, and analyzes EULA documents. These documents can either be uploaded or retrieved from a publicly accessible URL. The tool evaluates EULAs on a set of heuristics defined within the project.

If deploying to a debian/Ubuntu server environment, you may follow the quickstart guide to quickly install and initialize the program.

You will need the following things properly installed on your computer.

• A Unix or Linux based operating system
• Git
• Conda 2.7 for development or Native Pip if deploying to production.
• Node.js (with NPM)
• Java 8
• nginx
• uwsgi
• mongodb-dameon
• Google Chrome
• The Google chromedriver for your system, and available in your path.
• The C++ compiler for your system. For MacOS, this is included in XCode via xcode-select --install. For Linux systems, the package is g++ for Debian systems and gcc-c++ for Fedora systems.

• git clone https://github.com/REUL-Lab/eula-aat.git this repository
• cd eula-aat
• If developing, initialize and acgivate the environment with conda using
  • conda env create -f environment.yml
  • source activate eula-aat
• If deploying to production (on debian distributions only), run the quick configure script and skip to Part 2
  • sudo chmod +x setup.sh && sudo ./setup.sh
• For deploys to other environments, install the python environment and continue following
  • pip install -r requirements.txt

Install the punkt package for nltk

• python -c 'import nltk; nltk.download("punkt")'

Next, you must install python-boilerpipe. Be sure to do this in your home directory, not the project directory.

• cd /tmp
• git clone https://github.com/misja/python-boilerpipe.git
• cd python-boilerpipe
• pip install -r requirements.txt
• python setup.py install this.rootURL Once the installation is done, you may delete the python-boilerpipe directory - it is no longer needed.
• rm -rf /tmp/python-boilerpipe

• cd app
• npm install
  • Note that by default, node.js will attempt to allocate 4GB of memory. If you machine has less memory, use the command node --max-old-space-size=XXX /usr/bin/npm install where XXX is memory in MB.

Finally, run the nginx and uwsgi config script by typing ./configure.sh while in the project directory. This will append two lines to your .bashrc file to set the analyze_max_threads and google_api_key environment variables.

Note: Choosing the "test" option will just proxy requests from nginx onto your flask and ember debug systems. They must still be running for the request to serve properly. Choosing production will cause nginx to serve the requests itself.

If you wish to run the project immediately, run source ~/.bashrc to export the new environment variables set in the configuration script.

Run the flask service by activating the eula-aat environment as described above then running

• python api/app.py from the root directory
• nginx
  • If you specified a different configuration name during configure.sh, choose it by adding -c yourconfig.conf.
  • When done, stop the nginx daemon by running nginx -s stop
• mongod --fork to start mongodb as a daemon
  • When done, stop the mongodb daemon by runnning mongod --shutdown

Run the ember service by navigating to the /app directory then running

• ember serve

A guide to adding heuristics can be found in DEVGUIDE.md.

If you are exporting the static build to another webservice or S3, first edit app/config/environment.js:16 and replace this.rootURL with the address of the API server. Additionally, remove lines 33-36 in your nginx.conf to prevent the server from attemtping to serve the static content itself.

After installing the application and running the ./configure.sh script for production, set your webserver firewall to accept requests on port 80.

Create a build of the Ember application for nginx to serve by running ./node_modules/ember-cli/bin/ember build

Run the following command to initialize nginx as a service so it will start with your server: - sudo systemctl enable nginx

Configure mongodb as a service: - sudo systemctl enable mongod

Configure uwsgi as a service: - sudo systemctl enable uwsgi

Your server should now be ready to serve requests.