Updated! How I left MuckRock out is beyond me. There may be more updates as appropriate.
If you’re new to programming, looking at what others have done is probably the best way to get your bearings. DocumentCloud is no exception. You asked for more, better API examples. We’re long overdue for a roundup of some of the great tools DocumentCloud users have built on our API or otherwise poked their heads under the hood.
Mitch Kotler, one of the folks making MuckRock happen, wrote up his experiences integrating DocumentCloud into their Django site. His post on how their integration works is an excellent detailed walk through.
Chicago Tribune’s Chris Groskopf published a tidy little command line tool for managing bulk uploads. Jacob Fenton, who has done some great work integrating DocumentCloud projects into Investigative Reporting Workshop pieces, published doc-wrangler, which will let you wrangle DocumentCloud projects from your own machine.
Ben Welsh and his colleagues at The Los Angeles Times Data Desk maintain a public copy of the python wrapper they use to work with our API, and the folks at Bay Citizen used that python wrapper to integrate DocumentCloud and Django, so they could manage documents right in their CMS instead of logging in to DocumentCloud’s web interface.
WNYC’s John Keefe took Chicago Tribune’s document viewer, reworked it to meet his needs and published his customized viewer along with great documentation on integrating it into your website. NPR’s Chris Amico wrote a viewer in Flask while he was still at PBS NewsHour — his is another great starting point if you want to do more to customize document displays.
No Code, Still Worth Checking Out
As a bonus, if you’re on the lookout for great ways to incorporate documents into your reporting, definitely take a look at the San Antonio Reporter’s reporting on deed restrictions. They’ve assembled dozens of deeds and mapped them out for readers. From each map point, San Antonio Reporter readers can jump straight to the terms of the deed that has left a landlord with unrentable retail space.
The New York Times uses our API to offer readers a customized search of their Guantanamo docket that looks and feels very much a part of the Times’ website.
Likewise, The Telegraph built a custom search of the articles they’ve dug out of the News of the World archive where hacked voicemail or email may have contributed to reporting.
That can’t be all!
Did you publish some code of your own that builds on DocumentCloud’s API or the Times’ document viewer? We’d love to know about it!