Supreme Court of the United States : opinions

web citation discovery, presentation, and validation

https://scotus.law.berkeley.edu/

Setup

I recommend using python 3.8 to run this application. You will also need the standard python development libraries installed. Out of the box, this application depends on a mysql backend, and the instructions below cater to that dependency. You can easily use another database backend, see the Django documentation for instructions. Likewise, to run the application in production, you should also install and configure a web server, such a nginx or apache. Lastly, if you are installing this application on a server or VM where other websties, webapplications, or processes unrelated to this project are running, you should consider using virtualenv and virtualenvwrapper

To follow the directions below, you will also need to install pip, which is used to fetch the application requirements. Once that's taken care of, and mysql is installed and configured with a root user account, move into the directory where you want to store this application and then run the following:

git clone git@github.com:arderyp/scotuswebcites.git
cd scotuswebcites
./manage.py collectstatic
./setup.py

To run the opinion and web citation discovery, you can manually run the following from within the scotuswebcites/ directory:

./manage.py discover

However, it is more practical to have this process run automatically in your server's crontab. If you installed this application in your home directory on a unix OS without a virtualenv, and your user name is 'abc', you could simply add the following to your crontab: /home/abc/scotuswebcites/manage.py discover. If you are using virtualenv/virtualenvwrapper, you can modify then call the provided discovery script from your cron instead:

cp scripts/discover.sh.dist scripts/discover.sh
chmod 755 scripts/discover.sh

See the discover.sh file that you created for further instructions on using the script. For tips on setting up a unix cronjob, see this documentation.

Finally, you have the option to set up on-demand capturing via either webcitation.org, perma.cc, or both! What this means is that, once a user validates a citation, if that citation is a non-404, the application will utilize the webcitations.org/perma.cc APIs to grab an immediate capture of the citation. The links for these captures will be stored in your database and presented in the applications /citations/ view. To utlize the webcitation.org API, all you will need is a valid email address to get notifications. To utilize the perma.cc API, you will need to set up a perma.cc account. Once that is set up, you will need to log in to perma.cc, click your name in the upper right hand corner, click tools, then click "Generate API key". You should then contact perma.cc and ask them to create a shared folder for you to use with this application, then make note of the folder_id. You can activate the webcitations.org and/or perma.cc on-demand archiving by again editing the settings.py file. Run vi scotuswebcites/settings.py and find the following block:

WEBCITE = {
    'enabled': False,
    'api_query': 'http://www.webcitation.org/archive?returnxml=true&url=%s&email=%s',
}
PERMA = {
    'enabled': False,
    'archive_base': 'https://perma.cc',
    'api_key': 'PERMA_CC_API_KEY',
    'api_query': 'https://api.perma.cc/v1/archives/?api_key=%s'
}

Simply change False to True for the services that you want to enable. If you'd like to enable perma.cc, you will need to put the API key that you generated in place of PERMA_CC_API_KEY, and your shared folder_id in place of PERMA_CC_SHARED_FOLDER_ID. If you don't know what these are, contact someone at perma.cc for assistance.

Whether or not you have enabled archiving, you will need to validate citatons that are discovered. The scraper and the Opinion PDFs are imperfect, due to the methods by which opinion PDFs are generated, which means that some web citations are scraped improperly. Consequently, newly discovered citations need to be validated by a human to ensure that they are accurate. That being said, most scraped citations seem to be accurate these days. Once the citation is verified, if you have archiving enabled, the archives will be automatically generated and integrated. If you have enabled the gmail notification system via setup.py, you can now notify your subscribers of newly verified citations by clicking the "Nofity Subscribers" button at the top of the citations view. Likewise, you can run this opperation manually:

manage.py notifysubscribers

You may also set up a cron job to handle this automatically by utilizing scripts/notifysubscribers.sh.dist, following the same basic instructions provided for scripts/discovery.sh.dist noted above. If you choose to do so, you may want to stagger the discovery process to run every other hour, and have the notification process run every hour in between, which will give you (the admin) an hour to verify newly discovered citations before your runs the notification process. This is a long enough period to let you get your work done, but hopeuflly a short enough period that the cited website has not changed since your scraper discovered it. See the wiki for an example cron setup.

Lastly, all discovered opinion PDFs will be saved in the docs/pdfs directory for safe keeping. All discovery and notification logs are written to logs/, and each day gets it's own log file. That's it! If you encounter any bugs or experience issues setting up the application, feel free to open a ticket on this repository for further assistance.